# Agents API Overview Source: https://docs.nuwacom.ai/admin/api/agents-overview Create, manage, and chat with agents in your nuwacom workspace programmatically The Agents API lets you create, manage, and chat with agents in your nuwacom workspace programmatically, for example to provision agents from your own tooling, keep agent instructions in sync with an external source, or build custom chat experiences on top of them. ## Before you start * **API key**: All endpoints require a Bearer token. Admins can create API keys in the workspace settings under the **API Keys** section. * **Space scope**: Agents always belong to a space, so all endpoints are nested under the space path (`/api/v1/spaces/{spaceId}/agents`). The caller's role in that space must grant read permission for the Agents feature to list or read agents, and write permission to create, update, or delete them. * **Draft & published versions**: Create and update operations write to the agent's **draft**. [Publishing](/api-reference/agents/publish-an-agent) makes the draft the live version that the Completion API uses by default. API responses always show the draft state. The `published` field tells whether a published version exists, and `hasUnpublishedChanges` tells whether the draft differs from it. Who can see and use the agent is controlled separately via sharing in the nuwacom app. * **Custom agents only**: Built-in nuwacom system agents are not listed or readable through this API. Only agents created in the space (via this API or the app) appear. You can still chat with a system agent through the [Completion API](/api-reference/completion-api/openai-chat-completion) if you already know its ID. ## Base URL ``` https://{customer-tenant}.nuwacom.ai ``` Replace `{customer-tenant}` with your workspace's tenant name. ## Available endpoints ### Agents | Method | Endpoint | Description | | - | - | - | | `GET` | `/api/v1/spaces/{spaceId}/agents` | [List agents](/api-reference/agents/list-agents) | | `POST` | `/api/v1/spaces/{spaceId}/agents` | [Create a new agent](/api-reference/agents/create-an-agent) | | `GET` | `/api/v1/spaces/{spaceId}/agents/{agentId}` | [Get agent details](/api-reference/agents/get-an-agent) | | `PATCH` | `/api/v1/spaces/{spaceId}/agents/{agentId}` | [Update an agent](/api-reference/agents/update-an-agent) | | `POST` | `/api/v1/spaces/{spaceId}/agents/{agentId}/publish` | [Publish an agent](/api-reference/agents/publish-an-agent) | | `POST` | `/api/v1/spaces/{spaceId}/agents/{agentId}/unpublish` | [Unpublish an agent](/api-reference/agents/unpublish-an-agent) | | `DELETE` | `/api/v1/spaces/{spaceId}/agents/{agentId}` | [Delete an agent](/api-reference/agents/delete-an-agent) | ### Related | Method | Endpoint | Description | | - | - | - | | `POST` | `/api/v1/openai/chat/completions` | [Chat with an agent](/api-reference/completion-api/openai-chat-completion) via the `agentId` option | | `GET` | `/api/ai/models` | [List available models](/api-reference/ai/list-all-ai-models) | ## Pagination [List agents](/api-reference/agents/list-agents) is paginated. Control the page with two optional query parameters: | Parameter | Type | Description | | - | - | - | | `limit` | integer | Maximum number of agents to return (1–100). Defaults to `50`. | | `offset` | integer | Number of agents to skip from the start of the list. Defaults to `0`. | The response is an envelope: the agents are in `data`, alongside pagination metadata. ```json theme={null} { "data": [ { "id": "…", "name": "Support Agent", "published": true } ], "total": 128, "limit": 50, "offset": 0, "hasMore": true } ``` Page through the full list by increasing `offset` by `limit` until `hasMore` is `false`: ```bash theme={null} curl "https://{customer-tenant}.nuwacom.ai/api/v1/spaces/YOUR_SPACE_ID/agents?limit=50&offset=50" \ -H "Authorization: Bearer $NUWACOM_API_KEY" ``` ## Chat with an agent To chat with an agent, use the OpenAI-compatible [Completion API](/api-reference/completion-api/openai-chat-completion) and pass the agent's ID as `agentId`. The agent's full configuration (system prompt, knowledge sources, model options) is applied automatically: ```bash theme={null} curl https://{customer-tenant}.nuwacom.ai/api/v1/openai/chat/completions \ -H "Authorization: Bearer $NUWACOM_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "space_id": "YOUR_SPACE_ID", "agentId": "YOUR_AGENT_ID", "model": "azure-gpt-4o", "messages": [ { "role": "user", "content": "Hello!" } ] }' ``` The `model` field is required on the Completion API and acts as a fallback; when the agent has a `model` configured, the agent's model wins. By default, the **published** version of the agent is used. To chat with an agent that only exists as a draft (for example right after creating it via the API), additionally pass `"agentPublished": false`. ## Example: create an agent ```bash theme={null} curl -X POST https://{customer-tenant}.nuwacom.ai/api/v1/spaces/YOUR_SPACE_ID/agents \ -H "Authorization: Bearer $NUWACOM_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "name": "Support Agent", "description": "Answers customer support questions", "systemInstruction": "You are a friendly support agent. Answer concisely.", "conversationStarters": [ { "text": "What can you help me with?" } ], "model": "azure-gpt-4o", "temperature": 0.3, "attachmentIds": ["YOUR_ASSET_ID_AFTER_UPLOAD"] }' ``` Upload files first via [Upload an asset](/api-reference/assets/upload-an-asset), then pass the returned `id` in `attachmentIds`. Use `folderIds` or `contentIds` for knowledge folders and nuwacom documents. The available model IDs for `model` can be retrieved via [List available models](/api-reference/ai/list-all-ai-models). The response contains the new agent's `id` along with all its settings. The agent starts as an unpublished draft (`published: false`); use [Publish an agent](/api-reference/agents/publish-an-agent) to make it the live version. ## Actions Actions are the tools an agent may call while chatting — built-in nuwacom tools (like web search or knowledge retrieval), an [MCP](https://modelcontextprotocol.io) server's tools, or actions from a connected integration (Gmail, Outlook, …). Configure them via the `actions` array on [Create](/api-reference/agents/create-an-agent) and [Update](/api-reference/agents/update-an-agent). Each action has the following fields: | Field | Type | Description | | - | - | - | | `integration` | string | Provider the action belongs to: `"nuwacom"` for built-in tools, `"MCP"` for an MCP server, or a connected integration type (e.g. `"gmail"`). | | `key` | string | Action key within the integration. For `nuwacom`, one of the built-in keys listed below; for an integration or MCP server, the exact tool key it exposes (e.g. `"GMAIL_SEND_EMAIL"`). | | `requiresConfirmation` | boolean | Whether the user must confirm before the action runs. Defaults to `false`. | | `defaultValues` | object | Preset parameter values applied whenever the action runs. Defaults to `{}`. | | `integrationId` | string | ID of the specific connected integration instance to use (for provider integrations). Optional. | | `mcpClientId` | string | ID of the MCP client that exposes this action. Required for `"MCP"` actions. | ### Built-in nuwacom action keys When `integration` is `"nuwacom"`, `key` must be one of the app-supported built-in tools: | `key` | Description | | - | - | | `WEB_SEARCH` | Search the web. | | `RETRIEVE_FROM_KNOWLEDGE_BASE` | Retrieve from the agent's knowledge sources. | | `DISPLAY_DOCUMENT` | Render a document in the chat. | | `DISPLAY_SLIDES` | Render slides in the chat. | | `DISPLAY_EMAIL` | Render an email draft in the chat. | | `IMAGE_GENERATION` | Generate images. | | `VIDEO_GENERATION` | Generate videos. | > [!WARNING] > > Only the keys above are valid for `nuwacom` actions. Other internal nuwacom keys (e.g. `RETRIEVE_FROM_CONTENT`) are not configurable via the API — sending one is accepted by the API but is not a user-facing tool and will not render correctly in the app. For integration or MCP actions, use the exact key exposed by that provider/server. > [!NOTE] > > On update, the `actions` array **fully replaces** the agent's current draft actions. Send the complete desired set, or `[]` to remove all of them. Omitting `actions` leaves them unchanged. Like other changes, action edits apply to the draft — [publish](/api-reference/agents/publish-an-agent) the agent to make them take effect for the Completion API. ```bash theme={null} curl -X POST https://{customer-tenant}.nuwacom.ai/api/v1/spaces/YOUR_SPACE_ID/agents \ -H "Authorization: Bearer $NUWACOM_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "name": "Research Agent", "systemInstruction": "You research topics using web search.", "actions": [ { "integration": "nuwacom", "key": "WEB_SEARCH" }, { "integration": "nuwacom", "key": "RETRIEVE_FROM_KNOWLEDGE_BASE", "requiresConfirmation": true } ] }' ``` ## Skills Skills are reusable instruction packages (see the [Skills API](/admin/api/skills-overview)) that an agent loads on demand while chatting. Attach them via the `skillIds` array on [Create](/api-reference/agents/create-an-agent) and [Update](/api-reference/agents/update-an-agent); responses return the attached IDs in the same field. * When an agent has attached skills, **only** those skills are loaded for it. An agent without attached skills gets none. * Skill IDs must belong to the same space — create and discover them via [List skills](/api-reference/skills/list-skills) and [Create a skill](/api-reference/skills/create-a-skill). > [!NOTE] > > On update, the `skillIds` array **fully replaces** the agent's current draft skill set. Send the complete desired set, or `[]` to detach all skills. Omitting `skillIds` leaves them unchanged. Like other changes, skill edits apply to the draft — [publish](/api-reference/agents/publish-an-agent) the agent to make them take effect for the Completion API. ```bash theme={null} curl -X PATCH https://{customer-tenant}.nuwacom.ai/api/v1/spaces/YOUR_SPACE_ID/agents/YOUR_AGENT_ID \ -H "Authorization: Bearer $NUWACOM_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "skillIds": ["YOUR_SKILL_ID"] }' ``` ## Availability: app and external embed Where an agent can be used is controlled by two independent switches, both configurable on [Create](/api-reference/agents/create-an-agent) and [Update](/api-reference/agents/update-an-agent): | Field | Type | Description | | - | - | - | | `appEnabled` | boolean | Whether the agent is available inside the nuwacom app. Defaults to `true`. | | `embedEnabled` | boolean | Whether the agent is published as an external, embeddable chat widget. Defaults to `false`. | | `embedOptions` | object | External embed widget configuration (see below). On update the provided object **fully replaces** the current embed options. | | `embedId` | string | Read-only identifier used to embed the agent as an external widget. Returned in responses. | `embedOptions` supports the following keys: | Field | Type | Description | | - | - | - | | `domainWhitelist` | string\[] | Origins allowed to load the embed widget (e.g. `"https://www.example.com"`). Required to publish an embed-enabled agent. | | `showSources` | boolean | Whether the widget shows source citations. | | `allowFileUpload` | boolean | Whether end users can upload files in the widget. | | `allowVoiceInput` | boolean | Whether end users can use voice input in the widget. | | `themeMode` | `"light"`/`"dark"` | Default color theme of the widget. | | `lightTheme` | object | Light-theme colors (`backgroundColor`, `primaryColor` as HSVA objects). | | `darkTheme` | object | Dark-theme colors (`backgroundColor`, `primaryColor` as HSVA objects). | > [!NOTE] > > `appEnabled`, `embedEnabled`, and `embedOptions` are draft settings like everything else — [publish](/api-reference/agents/publish-an-agent) the agent to make them take effect. Publishing an agent with `embedEnabled: true` makes the external widget live and therefore requires a non-empty `embedOptions.domainWhitelist`; otherwise the publish request is rejected with `400`. ```bash theme={null} curl -X POST https://{customer-tenant}.nuwacom.ai/api/v1/spaces/YOUR_SPACE_ID/agents \ -H "Authorization: Bearer $NUWACOM_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "name": "Website Assistant", "systemInstruction": "You help visitors on our marketing site.", "appEnabled": false, "embedEnabled": true, "embedOptions": { "domainWhitelist": ["https://www.example.com"], "showSources": false, "themeMode": "light" } }' ``` # API Introduction Source: https://docs.nuwacom.ai/admin/api/introduction Integrate nuwacom's AI capabilities and your knowledge base into your own applications ## Overview The nuwacom API gives you programmatic access to your workspace: run AI completions with leading models, manage agents, and work with your knowledge base. Whether you are building custom applications, automating workflows, or enhancing existing systems, the API provides the building blocks you need. ## Available APIs ### Completion API Run AI completions through an OpenAI-compatible interface. Use plain models, enable knowledge base search and web search, or chat with a configured agent. * [OpenAI Chat Completion](/api-reference/completion-api/openai-chat-completion): OpenAI-compatible chat completions, including agent support via the `agentId` option ### Agents API Create and manage agents programmatically, with specialized instructions, conversation starters, and model options. * [Agents API Overview](/admin/api/agents-overview): Complete guide to working with agents * [List agents](/api-reference/agents/list-agents), [Create](/api-reference/agents/create-an-agent), [Get](/api-reference/agents/get-an-agent), [Update](/api-reference/agents/update-an-agent), [Delete](/api-reference/agents/delete-an-agent) ### Assets API Manage documents in your knowledge base for retrieval-augmented generation. * [Add an asset via URL](/api-reference/assets/add-an-asset-via-url) * [Upload an asset](/api-reference/assets/upload-an-asset) * [Get](/api-reference/assets/get-an-asset), [Update](/api-reference/assets/update-an-asset), [Delete](/api-reference/assets/delete-an-asset) ### AI Models Discover which models are available in your workspace and spaces. Use these model IDs in completion requests and agent options. * [List all AI models](/api-reference/ai/list-all-ai-models) * [List space AI models](/api-reference/ai/list-space-ai-models) ## Getting started ### Base URL Every nuwacom instance runs on its own tenant domain. Use your workspace's tenant name for all API requests: ``` https://{customer-tenant}.nuwacom.ai ``` ### Authentication All API requests require authentication using a Bearer token: ```bash theme={null} curl -H "Authorization: Bearer YOUR_API_KEY" \ "https://{customer-tenant}.nuwacom.ai/api/ai/models" ``` Admins can create and manage API keys in the workspace settings under the **API Keys** section. If you are not an admin, please approach your nuwacom admin to request an API key. > [!NOTE] > > Keep your API key in a secure backend environment. Never expose it in client-side code such as JavaScript running in a web page. ### Spaces Most resources (agents, assets, conversations) belong to a **space**. The corresponding requests take the space ID, either in the URL path (e.g. `/api/v1/spaces/{spaceId}/agents`) or as a `space_id` parameter, and the caller's role in that space determines what they are allowed to do. ## OpenAPI specification A comprehensive OpenAPI specification for your instance is available at `https://{customer-tenant}.nuwacom.ai/api/openapi.json` We provide two convenient interfaces to explore the OpenAPI documentation: * Scalar: `https://{customer-tenant}.nuwacom.ai/api/scalar` * Swagger UI: `https://{customer-tenant}.nuwacom.ai/api/swagger` # Skills API Overview Source: https://docs.nuwacom.ai/admin/api/skills-overview Create and manage skills in your nuwacom workspace programmatically The Skills API lets you create and manage skills in your nuwacom workspace programmatically, for example to keep skill instructions in sync with an external source, provision skills from your own tooling, or discover the skill IDs needed to attach skills to agents. A skill is a reusable instruction package for the AI: a name, a short description, and markdown instructions the AI loads on demand when they are relevant to the task at hand. ## Before you start * **API key**: All endpoints require a Bearer token. Admins can create API keys in the workspace settings under the **API Keys** section. * **Space scope**: Skills always belong to a space, so all endpoints are nested under the space path (`/api/v1/spaces/{spaceId}/skills`). The caller's role in that space must grant read permission for the Skills feature to list or read skills, and write permission to create, update, or delete them. * **User-created skills only**: Built-in nuwacom skills and skill templates are not exposed through this API. The API returns and manages only user-created skills (created manually, via upload, GitHub import, or AI generation). * **No draft state**: Unlike agents, skills have no draft/published distinction — changes are live immediately. ## Base URL ``` https://{customer-tenant}.nuwacom.ai ``` Replace `{customer-tenant}` with your workspace's tenant name. ## Available endpoints | Method | Endpoint | Description | | - | - | - | | `GET` | `/api/v1/spaces/{spaceId}/skills` | [List skills](/api-reference/skills/list-skills) | | `POST` | `/api/v1/spaces/{spaceId}/skills` | [Create a new skill](/api-reference/skills/create-a-skill) | | `GET` | `/api/v1/spaces/{spaceId}/skills/{skillId}` | [Get skill details](/api-reference/skills/get-a-skill) | | `PATCH` | `/api/v1/spaces/{spaceId}/skills/{skillId}` | [Update a skill](/api-reference/skills/update-a-skill) | | `DELETE` | `/api/v1/spaces/{spaceId}/skills/{skillId}` | [Delete a skill](/api-reference/skills/delete-a-skill) | | `POST` | `/api/v1/spaces/{spaceId}/skills/{skillId}/enable` | [Enable a skill](/api-reference/skills/enable-a-skill) | | `POST` | `/api/v1/spaces/{spaceId}/skills/{skillId}/disable` | [Disable a skill](/api-reference/skills/disable-a-skill) | | `GET` | `/api/v1/spaces/{spaceId}/skills/{skillId}/files` | [List skill files](/api-reference/skills/list-skill-files) | | `GET` | `/api/v1/spaces/{spaceId}/skills/{skillId}/files/{path}` | [Get a skill file](/api-reference/skills/get-a-skill-file) | | `PUT` | `/api/v1/spaces/{spaceId}/skills/{skillId}/files/{path}` | [Create or replace a skill file](/api-reference/skills/create-or-replace-a-skill-file) | | `DELETE` | `/api/v1/spaces/{spaceId}/skills/{skillId}/files/{path}` | [Delete a skill file](/api-reference/skills/delete-a-skill-file) | ## Instructions and SKILL.md Under the hood a skill is stored as a `SKILL.md` file: a YAML frontmatter block with the skill's `name` and `description`, followed by the markdown instructions. The API abstracts this away: * The `instructions` field on [Create](/api-reference/skills/create-a-skill) and [Update](/api-reference/skills/update-a-skill) is the **body** of the `SKILL.md` — write plain markdown, without frontmatter. * The frontmatter is generated automatically from `name` and `description` and kept in sync when you change them. * [Get a skill](/api-reference/skills/get-a-skill) returns the current body as `instructions`. The list endpoint omits it for performance; fetch a single skill to read it. > [!NOTE] > > On update, `instructions` **fully replaces** the current body. Read the current value first via [Get a skill](/api-reference/skills/get-a-skill) if you want to append rather than replace. Omitting `instructions` keeps the current body (a pure rename only updates the frontmatter). ## Activation Every skill carries an `enabled` flag in its API responses: the **space-wide** activation state. Enabled skills are loaded by the AI whenever their description matches the task; disabled skills are ignored. Skills start out enabled. * [Disable a skill](/api-reference/skills/disable-a-skill) switches a skill off for the whole space without deleting it — the skill and its files stay untouched. * [Enable a skill](/api-reference/skills/enable-a-skill) switches it back on. A space can have at most **40 active skills**; enabling beyond that limit returns `409` — disable another skill first. > [!NOTE] > > The API manages the space-wide state (admin semantics — API keys have no personal user context). Individual users can still override it for themselves in the nuwacom app; those personal preferences are not visible through this API. ## Skill files Besides its `SKILL.md`, a skill can carry additional files — reference documents, scripts, or templates the AI reads on demand when the instructions point to them. The file endpoints manage these files: * [List skill files](/api-reference/skills/list-skill-files) returns every file with its `path`, `size` (bytes), and `mimeType`. * [Get a skill file](/api-reference/skills/get-a-skill-file) returns a file's text content. Reference files by their relative `path` within the skill folder — nested paths like `reference/format.md` go directly into the URL. * [Create or replace a skill file](/api-reference/skills/create-or-replace-a-skill-file) writes a text file. Parent folders are created implicitly by the path; the provided `content` fully replaces any existing file. * [Delete a skill file](/api-reference/skills/delete-a-skill-file) removes a file. `SKILL.md` cannot be deleted — every skill needs one. **Text files only**: the file endpoints handle text formats (markdown, scripts, JSON/YAML, CSV, SVG, and similar). Binary assets such as images and fonts can currently only be uploaded in the nuwacom app. > [!NOTE] > > You *can* write `SKILL.md` through the file endpoints, which replaces the whole file including its frontmatter (a valid frontmatter `description` is synced back into the skill's metadata). Prefer [Update a skill](/api-reference/skills/update-a-skill) with `instructions` instead — it keeps the frontmatter consistent automatically. ## Pagination [List skills](/api-reference/skills/list-skills) is paginated. Control the page with two optional query parameters: | Parameter | Type | Description | | - | - | - | | `limit` | integer | Maximum number of skills to return (1–100). Defaults to `50`. | | `offset` | integer | Number of skills to skip from the start of the list. Defaults to `0`. | The response is an envelope: the skills are in `data`, alongside pagination metadata (`total`, `limit`, `offset`, `hasMore`). Page through the full list by increasing `offset` by `limit` until `hasMore` is `false`. ## Example: create a skill ```bash theme={null} curl -X POST https://{customer-tenant}.nuwacom.ai/api/v1/spaces/YOUR_SPACE_ID/skills \ -H "Authorization: Bearer $NUWACOM_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "name": "Meeting Notes", "description": "Formats meeting notes in the company template", "instructions": "When asked to write meeting notes:\n\n1. Start with date, participants, and agenda.\n2. Summarize decisions as bullet points.\n3. End with an action-item table (owner, task, due date)." }' ``` The response contains the new skill's `id`. The skill is active immediately and the AI can load it whenever its description matches the task. ## Field limits | Field | Limit | | - | - | | `name` | 1–64 characters; letters, digits, spaces, and hyphens; must start and end with a letter or digit. | | `description` | 1–1,024 characters. | | `instructions` | Up to 50,000 characters of markdown. | ## Use skills with agents Agents can have specific skills attached: pass the skill `id`s returned by this API in the `skillIds` array when [creating](/api-reference/agents/create-an-agent) or [updating](/api-reference/agents/update-an-agent) an agent. See the [Agents API overview](/admin/api/agents-overview#skills) for the full semantics (exclusive loading, full replace, draft/publish). # Advanced Features Source: https://docs.nuwacom.ai/advanced-features ## Using Fonts in Skills Fonts can be bundled directly into a Skill so that they are applied consistently whenever the Skill generates Word documents or templates. The requirements below determine whether a font is detected and embedded correctly, and they explain what happens when a font does not qualify. ### Folder Structure Font files are detected when they are placed in the assets/fonts/ folder inside the Skill. The following paths are also recognized: fonts/, images/, logos/, and img/. Fonts placed at the Skill root are not detected. ### Supported Formats TTF and OTF are the recommended formats. WOFF and WOFF2 are supported as well. Variable fonts cannot be embedded in Office documents and are therefore not accepted. ### Font Faces per Family Word supports four faces per font family: regular, bold, italic, and bold italic. Additional weights such as Light, SemiBold, or Black are folded into the base family, and only one weight per slot is used. To make a specific weight selectable in Word, ship it under its own distinct family name. ### Naming Convention Use the format FamilyName-Style.ttf, for example Montserrat-Bold.ttf. Numeric suffixes in the filename are also handled. ### License Requirements Fonts with an embedding restriction in the OS/2 fsType field are refused by design. When a font cannot be embedded, the document is generated with a fallback font instead. Verify that the font license permits document embedding before adding the font to a Skill. *** ## Orchestrator Skill Access The [Orchestrator](https://docs.nuwacom.ai/en/orchestrator) coordinates Agents, tools, models, and context to determine which AI capabilities are used, when they are used, and how they work together. It has access to the broadest range of Skills across all tiers. ### Skill Availability The following table shows which Skill types the Orchestrator can access: | Skill Type | Available to Orchestrator? | | :- | :- | | System skills (background) | Yes, always | | Workspace skills (toggled on) | Yes | | Personal skills (toggled on) | Yes | | Workspace or personal skills toggled off | No | The Orchestrator draws on all System Skills automatically, plus any workspace or personal Skills that are active at the time of the interaction. # Create an agent Source: https://docs.nuwacom.ai/api-reference/agents/create-an-agent /openapi.yaml post /api/v1/spaces/{spaceId}/agents Create a new agent in a space. The agent starts as an unpublished draft (`published: false`); publish it in the app to share it with the space. # Delete an agent Source: https://docs.nuwacom.ai/api-reference/agents/delete-an-agent /openapi.yaml delete /api/v1/spaces/{spaceId}/agents/{agentId} Permanently delete an agent, including its draft and published settings, actions and skills. # Get an agent Source: https://docs.nuwacom.ai/api-reference/agents/get-an-agent /openapi.yaml get /api/v1/spaces/{spaceId}/agents/{agentId} Return a single agent. Built-in nuwacom system agents are not available through this endpoint (404). The returned fields reflect the draft state (the version this API edits); `published` tells whether a published version exists. # List agents Source: https://docs.nuwacom.ai/api-reference/agents/list-agents /openapi.yaml get /api/v1/spaces/{spaceId}/agents List the agents of a space, ordered by most recently updated first. Built-in nuwacom system agents are excluded — only agents created in the space (via this API or the app) are returned. Results are paginated via the `limit` and `offset` query parameters; the response wraps the agents in `data` alongside `total`, `limit`, `offset`, and `hasMore`. The returned fields reflect the draft state (the version this API edits); `published` tells whether a published version exists. Use the agent ID with the [Completion API](/api-reference/completion-api/openai-chat-completion) (`agentId` option) to chat with an agent. # Publish an agent Source: https://docs.nuwacom.ai/api-reference/agents/publish-an-agent /openapi.yaml post /api/v1/spaces/{spaceId}/agents/{agentId}/publish Make the draft the published version of the agent. Copies the draft settings (including actions and skills) over the published version, so the Completion API uses them by default. If the draft has `embedEnabled: true`, publishing also makes the external embed widget live; this requires a non-empty `embedOptions.domainWhitelist`. Who can see and use the agent in-app is controlled separately via sharing in the nuwacom app. Idempotent: publishing an already published agent without draft changes is a no-op. # Unpublish an agent Source: https://docs.nuwacom.ai/api-reference/agents/unpublish-an-agent /openapi.yaml post /api/v1/spaces/{spaceId}/agents/{agentId}/unpublish Remove the published version of the agent; it reverts to a draft. The draft keeps all settings and can be re-published at any time. Published copies of actions and skills are removed along with the published version. Idempotent: unpublishing an unpublished agent is a no-op. # Update an agent Source: https://docs.nuwacom.ai/api-reference/agents/update-an-agent /openapi.yaml patch /api/v1/spaces/{spaceId}/agents/{agentId} Update an agent. Changes are written to the draft; the published version stays untouched until the agent is re-published in the app (`hasUnpublishedChanges` flags the drift). Omitted fields keep their current value. # List all AI models Source: https://docs.nuwacom.ai/api-reference/ai/list-all-ai-models /openapi.yaml get /api/ai/models List all supported AI models # List space AI models Source: https://docs.nuwacom.ai/api-reference/ai/list-space-ai-models /openapi.yaml get /api/ai/models/{spaceId} List all supported AI models for a specific space # Add an asset via URL Source: https://docs.nuwacom.ai/api-reference/assets/add-an-asset-via-url /openapi.yaml post /api/kb/assets Add an asset to the system by providing a URL to the asset # Delete an asset Source: https://docs.nuwacom.ai/api-reference/assets/delete-an-asset /openapi.yaml delete /api/kb/assets/{spaceId}/{assetId} Permanently delete an asset from the given space. This removes the stored file and any derived embeddings for that asset. # Get an asset Source: https://docs.nuwacom.ai/api-reference/assets/get-an-asset /openapi.yaml get /api/kb/assets/{spaceId}/{assetId} Return the complete asset record (name, MIME type, connector info, metadata, processing state, etc.) for the provided asset as long as the caller has read access to the space. # Update an asset Source: https://docs.nuwacom.ai/api-reference/assets/update-an-asset /openapi.yaml patch /api/kb/assets/{spaceId}/{assetId} Update mutable asset metadata (name, connector assignment, metadata JSON, file path info, etc.) without re-uploading the file itself. # Upload an asset Source: https://docs.nuwacom.ai/api-reference/assets/upload-an-asset /openapi.yaml post /api/kb/assets/upload Upload an asset to the system # OpenAI Chat Completion Source: https://docs.nuwacom.ai/api-reference/completion-api/openai-chat-completion /openapi.yaml post /api/v1/openai/chat/completions Create a new AI completion. You can find the supported models using [List all AI models](/api-reference/ai/list-all-ai-models). # Create a skill Source: https://docs.nuwacom.ai/api-reference/skills/create-a-skill /openapi.yaml post /api/v1/spaces/{spaceId}/skills Create a new skill in a space. The `instructions` become the body of the skill's SKILL.md file (its frontmatter is generated from `name` and `description`). The skill is active immediately; attach it to agents via the Agents API or in the nuwacom app. # Create or replace a skill file Source: https://docs.nuwacom.ai/api-reference/skills/create-or-replace-a-skill-file /openapi.yaml put /api/v1/spaces/{spaceId}/skills/{skillId}/files/{path} Create a new text file in the skill's folder or fully replace an existing one. Parent folders are created implicitly by the path. Writing to `SKILL.md` replaces the whole file including its frontmatter; when the new frontmatter contains a valid `description`, the skill's metadata is synced to it. Prefer [Update a skill](/api-reference/skills/update-a-skill) with `instructions` for SKILL.md changes — it keeps the frontmatter consistent automatically. Only text files can be written; binary assets are not supported via the API. # Delete a skill Source: https://docs.nuwacom.ai/api-reference/skills/delete-a-skill /openapi.yaml delete /api/v1/spaces/{spaceId}/skills/{skillId} Permanently delete a skill, including its SKILL.md and any other files. Agents that had the skill attached simply lose it. Built-in nuwacom skills cannot be deleted. # Delete a skill file Source: https://docs.nuwacom.ai/api-reference/skills/delete-a-skill-file /openapi.yaml delete /api/v1/spaces/{spaceId}/skills/{skillId}/files/{path} Delete a file from the skill's folder. `SKILL.md` cannot be deleted — every skill needs one; change it via [Update a skill](/api-reference/skills/update-a-skill) instead. # Disable a skill Source: https://docs.nuwacom.ai/api-reference/skills/disable-a-skill /openapi.yaml post /api/v1/spaces/{spaceId}/skills/{skillId}/disable Disable a skill for the whole space: the AI stops loading it for everyone (users can still re-enable it for themselves in the nuwacom app). The skill and its files stay untouched — use this instead of deleting when you want to switch a skill off temporarily. Re-activate it via [Enable a skill](/api-reference/skills/enable-a-skill). # Enable a skill Source: https://docs.nuwacom.ai/api-reference/skills/enable-a-skill /openapi.yaml post /api/v1/spaces/{spaceId}/skills/{skillId}/enable Enable a skill for the whole space, so the AI loads it whenever its description matches the task. Skills are enabled by default when created; use this to revert a [Disable](/api-reference/skills/disable-a-skill). A space can have at most 40 active skills — enabling beyond that returns 409. Individual users can still override the space-wide state for themselves in the nuwacom app. # Get a skill Source: https://docs.nuwacom.ai/api-reference/skills/get-a-skill /openapi.yaml get /api/v1/spaces/{spaceId}/skills/{skillId} Return a single skill, including its `instructions` (the body of the skill's SKILL.md file). # Get a skill file Source: https://docs.nuwacom.ai/api-reference/skills/get-a-skill-file /openapi.yaml get /api/v1/spaces/{spaceId}/skills/{skillId}/files/{path} Return the text content of a file in the skill's folder. Only text files can be read through this endpoint; binary assets (images, fonts) are not available via the API. # List skill files Source: https://docs.nuwacom.ai/api-reference/skills/list-skill-files /openapi.yaml get /api/v1/spaces/{spaceId}/skills/{skillId}/files List all files in the skill's folder, including its SKILL.md. Use the returned `path` values with [Get a skill file](/api-reference/skills/get-a-skill-file). # List skills Source: https://docs.nuwacom.ai/api-reference/skills/list-skills /openapi.yaml get /api/v1/spaces/{spaceId}/skills List the skills of a space, ordered by most recently updated first. Results are paginated via the `limit` and `offset` query parameters; the response wraps the skills in `data` alongside `total`, `limit`, `offset`, and `hasMore`. Only user-created skills are returned; built-in nuwacom skills are not exposed through this API. Use [Get a skill](/api-reference/skills/get-a-skill) to also read the skill's `instructions`. # Update a skill Source: https://docs.nuwacom.ai/api-reference/skills/update-a-skill /openapi.yaml patch /api/v1/spaces/{spaceId}/skills/{skillId} Update a skill. Omitted fields keep their current value. When `instructions` is provided it fully replaces the body of the skill's SKILL.md file; name/description changes are also synced into the SKILL.md frontmatter. Built-in nuwacom skills cannot be updated. # Account Source: https://docs.nuwacom.ai/en/account In the account settings section, you can manage your profile and customize your individual preferences for using your nuwacom workspace. To access, click your account in the bottom right of the navigation bar and select Account Settings. ![Account.png](https://mintcdn.com/nuwacomgmbh/QmokFyROSwNjqeyr/images/Account.png?fit=max\&auto=format\&n=QmokFyROSwNjqeyr\&q=85\&s=ba38e22161840d1d2ecf4691720e697f) *** ## Profile In the Profile section, you can view and update your personal information: * **Change or remove profile picture**: Click the pencil icon in the image area to upload a new picture or remove the existing one. * **First & Last Name**: Edit your personal details directly in the form field. * **Email Address**: For security reasons, your email address cannot be changed. *** ## Preferences Customize the basic appearance of your user interface and default AI model: * **Theme**: Choose between light, dark, or system-based appearance. * **Language**: Set the interface language (e.g., German, English or French). * **Preferred AI Model**: Select the AI model to be used by default in new chats (e.g., GPT-4.1). Learn more about AI models [here](/resources/aimodels). *** ## Memories ![Group 14713](https://mintcdn.com/nuwacomgmbh/osV_aASm-7aDJPwy/images/Group-14713.png?fit=max\&auto=format\&n=osV_aASm-7aDJPwy\&q=85\&s=6b845b2171997678e4dbbfedc7826afa) With memories, nuwacom stores important details from personal chats and takes them into account in future responses. This means recurring context does not need to be explained again every time. ### Enable memories 1. Open the profile settings. 2. Go to Memories. 3. Enable 'Remember content from chats'. When the feature is active, nuwacom can store helpful information from personal chats. ### Manage memories Click Manage memories to view, edit, or delete saved memories. New memories can also be added through an instruction in the AI chat, for example: ```text theme={null} Remember that I prefer short answers. ``` Memories remain private. Only the respective person has access to them. ### Import memories Memories from other AI tools can be imported. How it works: 1. Click Import memories. 2. Copy the displayed prompt. 3. Paste it into the other AI provider. 4. Copy the result. 5. Paste it into nuwacom. 6. Click Add to memories. ### FAQ > **When are chat memories used?** > > Memories are only used if you have enabled them in your personal settings. > **Where are saved memories applied?** > > Saved memories are used exclusively in personal chat conversations. They help nuwacom better adapt future responses to your preferences and work context. > **Are memories also used with agents or in the knowledge base?** > > No. Memories are not applied to agents or knowledge base content. They are used exclusively in personal chat conversations. > **Where can I manage my saved memories?** > > Under Profile settings > Settings, you can see all privately saved memories from your conversations. You can delete individual memories or remove all saved memories completely. > **Can I disable chat memories?** > > Yes. You can enable or disable chat memories at any time in your personal settings. ## Orchestrator The Orchestrator is the central system in nuwacom for handling complex tasks. It automatically coordinates multiple specialized agents and tools and can be activated via the model selection in the chat. In the settings, you define which agents are available for orchestration. Select the agents that should support complex task execution and activate them accordingly. This gives you control over which capabilities are included and how tasks are handled. *** ## Voices This feature allows you to analyze, define, and save specific tone-of-voice and writing styles, making them available across your entire workspace. All team members can easily apply these styles when creating content — ensuring consistent and brand-compliant communication throughout the organization. You have two options for creating a voice style: * **Identify a Voice Style (recommended)**: Let the AI analyze your writing based on samples or web pages and define the tone accordingly. You can manually refine the results afterward if needed. * **Define a Voice Style manually**: Describe the desired tone by considering characteristics such as formal or informal, emotional or objective, subjective or neutral, and consistent or dynamic. *** **Note:** For details on managing workspace settings and user permissions, please refer to the [**🔗 workspace settings article**](/product/workspacesettings) and [**🔗 user management article**](/product/usermanagement). # Agent Advanced Features Source: https://docs.nuwacom.ai/en/agent-advanced-features Beyond core configuration, the Agent Builder provides several features to organize agents across your workspace and improve the user experience when starting a conversation. *** ## Categories Categories are tags that help organize and find agents. Assigning a category to an agent makes it easier to understand its purpose at a glance and filter the agent library efficiently. ### Applying a Category to an Agent A category is set during agent creation or at any point in edit mode: 1. Open the agent in edit mode. 2. Navigate to the identity section (Avatar, Name, Description, and Category). 3. Select a category that matches the agent's purpose. The category is saved immediately and appears on the agent card in the library. ### Filtering by Category When browsing agents, the category appears on each agent card. Use it to quickly identify agents by domain or function, particularly in workspaces with a large number of agents. *** ## Pinning Agents Pinned agents appear in the sidebar for quick access. This is useful for agents used frequently, so they are always reachable without searching. ### How to Pin an Agent 1. Find the agent in the agent library or search results. 2. Open the three-dot menu on the agent card. 3. Select **Pin to sidebar**. To unpin an agent, open the same menu and select **Unpin from sidebar**. ### Accessing Pinned Agents Pinned agents appear directly in the navigation sidebar. Pinning is personal to each user account and does not affect how the agent appears to other workspace members. *** ## Agent Duplication Duplicating an agent creates an independent copy that can be modified without affecting the original. This is useful for creating variations of an existing agent or using another agent as a starting point. ### How to Duplicate an Agent 1. Find the agent in the agent library. 2. Open the three-dot menu on the agent card. 3. Select **Duplicate**. The copy opens in edit mode. The duplicating user becomes the owner of the new agent. ### What Gets Copied When an agent is duplicated, the following configuration is carried over to the new copy: * Name (with "Copy" appended) * Instructions * Avatar, description, and category * Actions and connected integrations * Knowledge sources * Agent start options, including conversation starters and welcome message * Settings (AI model, language style, response style) ### What Does Not Get Copied * Sharing settings: the duplicate starts as a private agent owned by the duplicating user * Usage statistics: the copy starts with no usage history *** ## Conversation Starters Conversation starters are pre-defined prompts that appear when a user opens an agent. They guide users toward effective interactions and help showcase what the agent can do. ### What They Are Conversation starters appear as clickable prompts on the agent start screen. Instead of typing a message from scratch, users select one of these prompts to begin the conversation. Up to 4 conversation starters are displayed at a time, selected randomly from the configured set. They are particularly useful for: * Demonstrating the agent's capabilities to new users * Guiding users toward common and high-value use cases * Reducing friction at the start of a conversation * Ensuring consistent, well-structured initial prompts ### How to Configure Them Conversation starters are configured in the Agent Builder under the agent start options: 1. Open the agent in edit mode. 2. Navigate to the start options section. 3. Select **Conversation Starters** as the start type. 4. Add starter prompts via **+ Add Prompt**. A welcome message can also be configured here. ### Best Practices * Write starters that reflect specific, realistic use cases rather than generic prompts. * Cover a range of tasks the agent handles, not just the most common one. * Keep starters concise so they are easy to read and select. * Update starters over time based on how users actually interact with the agent. # Agent Analytics Source: https://docs.nuwacom.ai/en/agent-analytics ## Overview Analytics provides workspace administrators with aggregated insights into how the workspace is used. Within Analytics, the Agents tab brings together all metrics that show how Agents are used within the selected date range. This makes it possible to review Agent usage, compare it across different periods, and prepare it for reporting. Evaluations follow the selected date range and grouping, for example by day. > [!NOTE] > > Agent Analytics is available to workspace administrators only. ## How to view Agent analytics 1. Open *Workspace settings*. 2. Select *Analytics*. 3. Choose the *Agents* tab. 4. Set the Date range and Group by filters. The filters above the view control how data appears across charts and metrics. When a filter changes, all affected charts and metrics update based on the selected settings. *Group by* defines the time interval used to aggregate the data, for example day, week, or month. *Date range* defines the period used for the evaluation, for example today, last week, or a custom range. ## Metrics in the Agents tab | Metric | Description | | :- | :- | | Agent messages | Shows the number of user messages sent in conversations with an Agent. | | Active Agent users | Shows the number of individual users who sent at least one message in an Agent conversation. | | Messages per Agent | Shows how many user messages were sent to each Agent. | | Agent model usage | Shows which AI models were used for answers in Agent conversations. | Agent filters are available for the *Active agent users* and *Messages per agent* metrics. Selecting one or more agents makes it possible to analyze how a specific combination of agents was used. Hovering over a chart bar displays the exact value for the selected point in time. Message-based metrics count user messages. Generated AI answers are reflected separately in Agent model usage. ## Exporting data Select **Export CSV** on a chart to download the data currently represented by that chart. Each export contains one row per data point, for example one row per model, and the corresponding aggregated count for the selected date range. The filename includes the chart type and the selected time period. The exported data can be processed outside nuwacom, for example in Excel, a reporting tool, or internal analyses. ## Additional metric information The information icon next to an evaluation opens a short description of the corresponding metric. This helps verify what a chart or metric represents before the data is used in reports. ## FAQ > **Who can access Agent analytics?** > > Agent analytics is available to workspace administrators only. Users without administrator permissions do not have access to the Analytics area in workspace settings. > **Why do Agent messages and Agent model usage show different numbers?** > > Message-based metrics such as *Agent messages* count user messages sent in Agent conversations. *Agent model usage* reflects the generated AI answers instead, which are tracked separately. As a result, the two metrics measure different events and can show different values. > **Does changing the date range or grouping affect all Agent charts?** > > Yes. The *Date range* and *Group by* filters apply to the entire *Agents* tab. When either filter changes, all affected charts and metrics update based on the selected settings. # Agent Builder Source: https://docs.nuwacom.ai/en/agent-builder ## What is the Agent Builder? The Agent Builder is where nuwacom agents are created and configured without code. An agent handles complex, recurring tasks from the initial request to a reviewed result using shared company context.Admins retain central control over permissions, approvals, and logging. Once an agent is saved, it can be called directly in the chat by mentioning it with "@". ## Before you start Defining how the agent should help produces a clearer structure from the outset. Four questions guide the setup: * **Goal**: What should the agent achieve for the user? * **Steps**: Which logical sub-steps make up the task? * **Knowledge**: Does the agent draw on a single file, multiple documents, or examples? * **Input type**: Do user inputs arrive as free text, structured forms, or targeted follow-up questions? ## Creating an agent with the builder A new agent is created via the Create Agent option in the Agents module. Three paths are available. ### Start with a template nuwacom provides predefined agents: a searchable collection of agent templates that cover proven use cases out of the box. Instead of starting from zero, a matching template is selected and then adapted to the brand and requirements. Templates are organized by business area so the right one is found quickly: * **Sales**: for example, Proposal Assistant, Competitive Intelligence Agent * **Customer Support**: for example, Knowledge Base Assistant, Customer Feedback Insights Agent, Support Triage Agent * **Marketing**: for example, Brand Review Agent, Content Repurposing Agent, Campaign Briefing Agent * **Legal**: for example, Data Privacy Review Agent, Contract Review Agent * **Product**: for example, Document Intelligence Agent, Product Research Agent, Release Notes Agent ### Create with AI Describing what the agent should do in the input field starts an AI-assisted build, and the agent is assembled automatically. This method saves time and helps with formulating the right instructions. Every detail can be fine-tuned manually afterward. ### Create from scratch The from-scratch option configures the agent step by step, with name, instructions, knowledge, and behavior defined manually. It is available via the three-dot menu and offers maximum control with specific details from the start. ## Refining and testing An agent is tested during creation and after configuration. The "Run" preview shows how the agent works and supports optimizing the instructions. Several iterations are normal. If something does not work as intended, a clearer description or a revision of the relevant section usually helps, and running typical use cases confirms that the instructions hold in different situations. ## Saving, sharing, and publishing After configuration, an agent stays private or can be shared with users, user groups, or the whole workspace. When sharing, choose whether recipients have read or edit rights. The share button makes the agent available in the workspace or shares it with specific people. > ***Important***\*: Knowledge connected to the agent is shared with those users as well. When an agent with attached prompts is shared, the prompts are shared too. When an agent with actions is shared, permissions also matter: users either set up their own accounts or use the existing account.\* > ***Tip***\*: After saving, an agent can be pinned to the navigation bar for quick access.\* ## FAQ > **When should a template be used instead of building from scratch?** > > A template provides a proven base structure that only needs to be adapted to the specific context, which suits a fast start. Building from scratch suits cases where the setup is already clear and maximum control is needed. > **Does the AI configure the entire agent automatically?** > > The AI assembles the agent from a description and can set most of the configuration. Every detail can then be refined manually in the editing area. > **How is an agent used after it is created?** > > Saved agents are called directly in the chat by mentioning them with "@". > **What happens to connected knowledge when an agent is shared?** > > Knowledge linked to the agent is shared together with the agent, so shared users gain access to that content as well. # Agent Configuration Source: https://docs.nuwacom.ai/en/agent-configuration The Agent Builder provides configuration options for adapting an agent to specific use cases. Instead of setting the options below manually, you can describe your agent in the Agent Builder chat and let it configure them for you. Both approaches work on the same configuration, so you can switch between them at any time. The following configuration options are available: ## Avatar, name, description and category An avatar gives the agent a visual identity. Name and short description are entered, unless the AI has already done this. An appropriate category ensures clear organization of the agents. *** ## Instructions Instructions define how the agent responds and which tasks it performs. Clear instructions enable effective work and the desired results. Two methods are available: **Generate instruction with AI**\ Using "Create" the AI generates detailed instructions. This saves time and helps with uncertain wording: based on a short description, clear, effective instructions are produced for reliable operation. **Write instructions manually**\ The instructions can also be defined manually. This option offers full control and flexibility to directly include specific requirements and details. Ideal for experienced users who know exactly how their agent should act. The choice between automatic generation and manual creation allows the method that best fits needs and level of knowledge, for optimal results in configuration. ### Structure of clear instructions Clear instructions arise from three key elements: **Task, Action and Goal**. Supplementary "Response Guidelines" control tone and writing style. Together they define what the agent does, how it responds and how the results are structured. **Task**\ The role describes who or what the agent should be. This influences tone, expertise and communication style. Example: Act as an automated messaging assistant for Slack that enables users to send messages to specific channels or people. **Action**\ The action describes the individual steps of the agent. This becomes especially important when multiple actions are involved. The clearer the steps, the better the result. Example, if users want to send a message: 1. Ask for the channel or recipient. 2. Display a preview of the message. 3. Send the message to the specified recipient. 4. Confirm the sending. **Response Guidelines**\ The response guidelines specify how the agent replies, for example regarding tone, writing style or approach. Example: * Use a friendly and concise language. * Ensure clarity when confirming recipient and message content. * Handle errors confidently and offer steps to resolve them if sending fails. * Keep a log of sent messages upon request. **Goal**\ The goal defines in one to two clear sentences what the agent should achieve for the user. Example: Optimize communication in Slack by providing a fast and reliable way to send messages, in order to increase team collaboration and productivity. *** ## Agent start options When creating, it is determined what happens when the agent is selected. This guides users from the start. Two options are available: **Start action**\ Automatically triggers a prompt when the agent is selected, also presented as a form to control user inputs. Ideal for structured, recurring processes. Use "+ Add prompt" to define the start form. **Conversation starters and welcome message**\ Instead of a fixed start action, optional conversation elements make the interaction more flexible and inviting: * **Conversation starters**: Up to 4 different conversation starters appear randomly. They inspire users and show possible questions or actions. Use "+ Add prompt" to define start prompts and a welcome message. * **Welcome message**: A short welcome message invites interaction. It appears as soon as the agent is selected but no action has yet started. Suitable for open, informal interactions without structured input. *** ## Actions ### Adding actions from different tools Actions bind tools like Slack, Google Drive, Microsoft Teams and others. This way agents not only respond but perform tasks: send messages, manage calendars or retrieve files. Complete workflows run directly in the chat.\ Supported tools include: ·       **Slack**: send messages, notify channels ·       **Microsoft Teams**: send messages, notify channels ·       **Gmail**: draft or send emails ·       **Google Calendar**: retrieve, add or update events ·       **Google Drive**: retrieve files, add documents ·       **SharePoint**: retrieve files, add files ·       **OneDrive**: manage documents Multiple connectors can be combined within one agent. [The full article shows all integrations and use cases.](https://docs.nuwacom.ai/de/produkt/konnektoren) ### **Adding nuwacom Actions** In addition to actions from connected tools, nuwacom-native actions can be added — for example, to enable web search or access to the Knowledge Base. **To add nuwacom actions:** 1. Open the agent and navigate to *Actions.* 2. Select *Add action.* 3. Choose *Nuwacom.* 4. Add individual actions as needed, or select *Add Nuwacom* to add all available nuwacom actions at once. **Available actions:** * **Search Knowledge Base**: Searches the Knowledge Base for relevant information. * **Search Web**: Searches the web for current information and answers. * **Write content**: Creates and edits content in editor. * **Create email**: Creates and edits emails. * **Create Slides**: Creates and edits presentations. * **Generate images**: Creates and edits images using AI. * **Generate videos**: Creates short videos based on text descriptions. Once added, the agent uses the selected actions automatically when they are relevant to a given request. *** ## Knowledge Knowledge enables more accurate, relevant and contextual responses, especially for company-specific information. **Example**: A support agent accesses stored product documentation and FAQ documents for each response and formulates customer replies solely based on these linked contents. Knowledge comes from the following sources: * **Upload files**: Upload new documents directly. * **Use "My files"**: Select already existing personal files. * **Include knowledge folders**: Add entire knowledge folders, individual subfolders or specific files. Changes in workspace folders are automatically applied to all linked agents. Manually updating individual agents is not required. *** ## Skills Skills can be attached directly to an Agent to define how it responds. Once a Skill is added, it is automatically active for everyone using that Agent. Users don’t need to install or select the Skill themselves.This means you can provide a fully configured Agent with tone, structure, and rules already built in. Users simply open the Agent and start working, while every output follows the same standard. **Example**\ An Agent with a *Meeting Notes Skill* always structures outputs into Key Decisions, Open Questions, and Next Steps — no setup required for the user. **Skills vs. Agents — The Critical Distinction** | | **Skill** | **Agent** | | :- | :- | :- | | **Defines** | How nuwacom responds | What nuwacom does | | **Focus** | Format, tone, structure | End-to-end task execution | | **Mode** | Passive — activates automatically or on demand | Active — can run autonomously | | **Think of it as** | The rulebook | The executor | Combining a Skill with an Agent enables autonomous task execution with consistent output quality. If your team has to navigate somewhere deliberately for something that should just always work, that behavior belongs in a Skill. *** ## Settings The "Settings" tab adjusts the agent's behavior and interaction and controls tone, intelligence and availability. **Suggest replies automatically**\ The AI suggests a reply after each agent message, promoting the flow of conversation. **AI model**\ A specific model ensures consistent capabilities or responses. \ Learn more about AI models here \{Link}. **Voice**\ A voice that matches the brand or preferred tone can be selected. **Response style**\ Determines how the agent replies: ·       **Precise**: for accurate, fact-based answers ·       **Creative**: for brainstorming or open tasks ·       **Balanced**: a mix of both *** ## Publish agent externally An agent can be embedded outside the workspace, for example on a website via widget or iFrame. 1. **Enable external publishing**: In the "Publish agent externally" area, turn the switch on. 2. **Set allowed domains**: Enter one or more domains into the whitelist to control embedding locations. Example: [https://yourcompany.com](https://yourcompany.com). Click "Save" to apply the domain restrictions. 3. **Choose embedding option**: Use iFrame or widget. **Option A: iFrame**\ Creates a fullscreen conversation experience, ideal for intensive interactions or support. After publishing, the iFrame code is copied and inserted into the website's HTML. **Option B: Widget**\ Adds a floating AI chatbot that provides quick help without disrupting the user experience. After publishing, the widget code is copied and placed before the page's closing tag. The embedding code is only available after the agent is published. **Test integration**\ Check on the website whether the agent loads correctly. Verify styling, expected response behavior and effective domain restrictions. If you have questions about embedding, the support team will be happy to help. *** ## Test agents An agent is tested during creation and after configuration. The "Run" preview shows how it works and serves to optimize the instructions. Multiple iterations are normal. If something doesn't work as desired, a clearer description or revision of the relevant part helps. Typical use cases ensure that the instructions apply in different situations. *** ## Save and share agent After publication the agent remains private or is shared. Use the share button to make it available in the workspace or share it with specific people. **Important:** Knowledge linked to this agent is also shared with the users. When sharing an agent with attached prompts, the prompts are also released. When sharing an agent with actions, permissions must also be considered: Do users set up their own accounts or do they use the existing account? # Agent Templates Source: https://docs.nuwacom.ai/en/agent-templates ## What Are Agent Templates Agent Templates are pre-built agent configurations that nuwacom provides as a browsable collection covering proven use cases. Each template ships with a complete structure of instructions, settings, skills, and actions, so a working agent is available immediately instead of being built from scratch. Templates are organized by business area, which makes it fast to locate a fit and then adapt it to your brand and requirements. Templates are best suited for a quick start: a matching template delivers a tested foundation that only needs to be aligned with your own context. Every building block remains fully configurable in the editor after creation. ## Browsing the Template Library The template library opens from Agents in the sidebar via the Browse Templates button. Each template is shown as a card with its name, a short description, and its creator. ### Getting Started The Getting Started templates are onboarding agents that help new users find their footing on the platform before building for a specific business area. Rather than solving a departmental task, they guide the choice of use case, answer platform questions, and improve how you prompt, so the first steps with agents are faster and more confident. Three are featured: * **Use Case Finder**: Helps you discover the best AI agent use cases for your team. * **nuwacom Expert**: Answers any question about the nuwacom platform and features. * **Prompt Optimizer**: Refines and improves your prompts for better agent performance. From the same page, Browse Templates opens the full collection of ready-made agents for Marketing, Legal, HR, Sales, and more. ### Agent Templates The library provides two ways to narrow down the selection: * **Search**: Use the Search templates field to find templates by name or description. * **Categories**: Filter by business area to focus on relevant templates. Available categories include Sales, Customer Support, Marketing, Legal, and Product. Representative templates per category: 1. **Sales**: Proposal Assistant, Competitive Intelligence Agent 2. **Customer Support**: Knowledge Base Assistant, Customer Feedback Insights Agent, Support Triage Agent 3. **Marketing**: Brand Review Agent, Content Repurposing Agent, Campaign Briefing Agent 4. **Legal**: Data Privacy Review Agent, Contract Review Agent 5. **Product**: Document Intelligence Agent, Product Research Agent, Release Notes Agent ## Template Details Selecting a template card opens the Create a New Agent preview. It displays the template name, an avatar, and a natural-language prompt field for describing changes, along with the full configuration under four building blocks. The Create Agent button adds a copy of this configuration to your workspace. The Brand Review Agent template illustrates how these building blocks fit together. ### Instruction The Instruction tab defines what the agent does and how it operates. In the Brand Review Agent, the instructions are structured into clear, named sections: * **Role and Purpose**: Establishes the agent's mandate, for example evaluating written content against brand standards across brand fit, tone of voice, writing quality, and terminology consistency before it goes out, surfacing issues and proposing fixes without publishing or making final decisions. * **Review Dimensions**: Breaks the task into concrete, numbered criteria the agent evaluates every time, such as Brand Fit, with each dimension pointing to the skill that carries its detailed logic. This structure shows the recommended pattern: a role that frames the agent's purpose, followed by explicit, rule-based instructions that delegate detailed logic to skills. ### Settings The Settings tab controls the agent's model behavior. In the Brand Review Agent, these are: * **Model**: The AI model powering the agent, for example Claude Sonnet 4.6. * **Temperature**: The value that governs how deterministic or creative responses are, for example 0.7. * **Follow-up messages**: A toggle that enables suggested follow-up messages, set to On. ### Skills The Skills tab lists the packaged knowledge and rule sets the agent draws on. Each skill appears as a card with a name and a description of when it applies. The Brand Review Agent bundles four: 1. **Brand Kit Guardian**: Ensures content aligns with brand standards, including positioning, tone of voice, visual and linguistic guardrails, and Dos & Don'ts. 2. **Tone of Voice & Communication Style**: Enforces rules for tonality, language register, phrasing, and communication style across written output. 3. **Writing Style Guide**: Defines rules for clarity, structure, formatting, headlines, lists, CTA style, and readability. 4. **Product Glossary**: Defines product terms, feature names, technical terms, and unified explanations for internal and external use. Skills keep detailed logic modular, so the instructions can orchestrate and consolidate findings while each skill supplies the specifics referenced by a review dimension. ### Actions The Actions tab connects the agent to tools so it can complete tasks, not only respond. The Brand Review Agent includes nuwacom actions: * **Write content**: Produces written output. * **Search Knowledge Base**: Retrieves relevant information from connected knowledge. Beyond the nuwacom actions shown here, templates can combine external connectors within a single agent, including Slack, Microsoft Teams, Gmail, Google Calendar, Google Drive, SharePoint, and OneDrive. ## How to create agent from a Template 1. Open the Agents module. 2. Select Browse Templates. 3. Choose a suitable template, for example "Brand Review Agent". 4. Review the Instruction, Settings, Skills, and Actions in the Create a New Agent preview. 5. Select Create Agent to add a copy to your workspace and open the editor. 6. Adjust the agent in the editor to match your requirements. 7. Test the agent in the Run preview. Refine the instructions or other settings as needed and repeat the test. 8. Save the agent. 9. Share the agent selectively with individual users, user groups, or the entire workspace when needed. **Tip:** After saving, an agent can be pinned to the navigation bar to keep it quickly accessible. **Important:** When an agent is shared, any knowledge connected to it is shared with those users as well. For agents with attached prompts or actions, review the prompts and the permission model, specifically whether users connect their own accounts or use the existing one. ## Customizing Your Copy The copy created from a template is fully independent and configurable in the agent configuration view \{Link}. Each building block maps to what you can change: * **Instruction**: Modify the role and review criteria to fit your process, or regenerate the instructions with AI from a short description in the prompt field. * **Settings**: Change the model, adjust the temperature, and toggle follow-up messages. * **Skills**: Add or remove skills, and attach knowledge by uploading files, selecting from My Files, or linking knowledge folders. Changes in workspace folders are applied automatically to all linked agents. * **Actions**: Add or remove nuwacom actions and external connectors. * **Identity and sharing**: Rename the agent, update its description and avatar, assign a category, and configure who can use it. ## FAQ > **When should I use an Agent Template?** > > Use a template when a proven configuration is a faster starting point than building an agent from scratch. Templates cover common use cases, but the instruction, settings, skills, and actions should still be reviewed before the agent is used with real workflows. > **What should I adjust after creating an agent from a template?** > > Adapt the agent to your own context: refine the instruction, confirm the settings, connect the relevant skills and actions, and test with realistic scenarios before sharing. > **Can I change the agent later?** > > Yes. The copy is independent from the template, so every building block remains editable in the editor at any time, and later changes to the original template do not affect your agent. # AI Usage Source: https://docs.nuwacom.ai/en/ai-usage ## Overview ![Aiusageenglish](https://mintcdn.com/nuwacomgmbh/FwrUFwR_NNOqGXf2/images/aiusageenglish.png?fit=max\&auto=format\&n=FwrUFwR_NNOqGXf2\&q=85\&s=0407e40068884cf682abfecbd90486aa) AI usage is now part of nuwacom and makes visible how much AI performance is used during the current usage period. The model creates transparency about consumption and flexibility in how AI performance is distributed across different Seats and expanded when needed. The included AI usage is part of every Seat and covers normal day-to-day work without requiring each action to be planned individually. When demand rises, Additional AI Usage can be booked and flexibly distributed across users. Three terms structure the model: * **Plans** define the feature scope and provisioning, bookable as Business or Enterprise. * **Seats** determine, via the Seat-Type, how much AI usage is included per user, available as Standard, Premium, and Power. * **Additional AI Usage** expands the available capacity at the company level and is assigned to individual users as needed. *** ## How AI Usage is structured ### AI Usage in Seats The included AI usage is covered through the Seat. Three Seat-Types are available: Standard, Premium, and Power. The Seat-Types differ only in the amount of included AI usage, not in feature scope: every user receives the same complete feature scope of the respective Plan. ### Additional AI Usage When the included AI usage is not sufficient, Additional AI Usage can be booked. It is distributed flexibly across users and suits higher or targeted demand, such as extensive workflows, building apps, or intensive agentic work. ### Separate Usage Areas Certain use cases run separately from the internal Seats, in particular external chatbots without an internal user license. Separate rules for Additional AI Usage apply to these areas, agreed upon based on expected usage, model choice, and use case. *** ## Seats at a Glance The Seat-Type determines how much AI usage is included per user and which price per user applies. The following overview shows which Seat-Type suits which users and what AI usage is included. Users who permanently need more AI usage move from the Standard Seat-Type to Premium, or from Premium to Power. The Premium Seat-Type closes the gap for active users who need more AI usage but do not yet require a Power Seat. | Seat | Suitable for | Included AI usage | Price per user / month | | :- | :- | :- | :- | | **Standard** | Most employees in normal day-to-day work: chat, knowledge search, summaries, meetings, simple documents | 1x | €25 | | **Premium** | Users who regularly work more intensively with AI: longer files, many meetings, frequent documents, first workflows, visualizations | 2x | €49 | | **Power** | Very active users, AI Champions, or Builders: many files, many tests, agents, complex tasks, and building apps and workflows | 5x | €99 | Users who permanently need more AI usage move from the Standard Seat-Type to Premium, or from Premium to Power. The Premium Seat-Type closes the gap for active users who need more AI usage but do not yet require a Power Seat. *** ## Managing AI Usage in the Workspace Admins manage AI usage centrally in the settings under Plan & Usage. This area is accessible to admins only. Plans, Seats, consumption, and Additional AI Usage are brought together there. > [!NOTE] > > For Enterprise plans, plan and billing management is handled through the nuwacom Customer Success Manager. ### View AI Usage ![Openaiusagesettings](https://mintcdn.com/nuwacomgmbh/8VOiVNfdltJo1V7h/images/openaiusagesettings.gif?s=90a7ebde160dd0a02ade26073f724e98) 1. Click on *Account*. 2. Select *View AI Usage.* 3. Review your teams AI usage. You can also access AI Usage centrally in Workspace Settings>Plan & Usage. ### Review Plan and Included Seats View the active Plan and renewal date, including the included Seats per Seat-Type: Standard (€25), Premium (€49), Power (€99). Unassigned Seats can be assigned directly through the displayed notice. Plan and billing can be adjusted through *Manage Plan* and *Payment and Billing*. ### Review Consumption in the Workspace View the consumption of all users at a glance. Track the remaining Additional AI Usage in total and per user as utilization. ### Assign Seats and Additional AI Usage For each user, see which Seat-Type is assigned, along with the assigned Additional AI Usage and utilization as a percentage. 1. Change the **Seat-Type** in the "Seat" column. 2. Assign **Additional AI Usage** from the workspace quota to individual users in the "Additional" column. 3. Add more users through "Invite Users" or filter through the search field. *** ## Providing more AI Usage Two paths increase the available AI usage: ### **1. Upgrade the Seat** Raise a user's Seat-Type, for example from Standard to Premium or from Premium to Power, when more AI usage is permanently needed. ![Manageplan](https://mintcdn.com/nuwacomgmbh/rNPBxkjulYS6_TD-/images/manageplan.gif?s=7700517aa473c380b2dc5e923e49b702) ### **2. Book Additional AI Usage** Rrovide more flexible capacity for users, teams, or use cases. ![Additionalaiusage](https://mintcdn.com/nuwacomgmbh/rNPBxkjulYS6_TD-/images/additionalaiusage.gif?s=f9a4d2eef4946ab32074497c67541bb0) Additional AI Usage is booked at the company level and can be distributed flexibly to the users, teams, or use cases that need it. It is especially suitable for more demanding or recurring usage, such as extensive workflows, building apps, or many automations. *** ## AI Usage Availability Period ### Seats The AI usage included with a Standard, Premium, or Power Seat is available for one **month** at a time. The full allowance is renewed at the beginning of each month. Unused included AI usage expires at the end of the month and does not carry over to the next month. Each month therefore starts with the full monthly allowance, regardless of how much was used in the previous month. ### Additional AI Usage Additional AI Usage is booked as an allowance for the respective **contract term**. This allowance is available throughout the entire contract term and can be used flexibly during this period. Additional costs only apply if the customer actively books Additional AI Usage or authorizes its purchase. *** ## Plans at a Glance nuwacom is bookable in two Plans. Both include the full AI functionality and differ in provisioning, governance, hosting, and integration depth. | Plan | Suitable for | Hosting | Pricing | | :- | :- | :- | :- | | **Business** | Teams that want to work with AI quickly | Europe on Azure | From €25 per user / month | | **Enterprise** | Companies with company-wide AI deployment and maximum control | Europe on Azure, optionally Germany on STACKIT | Individual, from 20 users | The Enterprise Plan includes all Business features and adds automated Knowledge Base synchronization, advanced governance and compliance, custom integrations, prioritized support with SLA, and dedicated onboarding. A detailed feature and pricing overview is available [here](https://nuwacom.com/en/pricing). To unlock advanced features, switch to Enterprise and [talk to an AI expert](https://insights.nuwacom.ai/meetings/robert-ivanov/demo?uuid=5b6d58fb-481d-46b1-8818-416c45ac9fb7). *** ## What happens when AI Usage is exhausted ### Normal Usage in nuwacom When the monthly included AI usage is used up, nuwacom does not block. Chat remains available. For further requests, the AI model automatically switches to the [Auto Mode Instant model](https://docs.nuwacom.ai/en/aimodels#auto-mode), allowing users to continue working without interruption or unplanned additional costs.\ > [!NOTE] > > See the active Instant model under *Workspace settings → Models*. Auto Mode selects the best model automatically. ### External Chatbots and Workflows When the Additional AI Usage designated for an external chatbot or workflow is exhausted, it is blocked. Further processing resumes only once Additional AI Usage has been booked, a Seat upgrade has taken place, or the next monthly usage period starts. *** ## What consumes more AI Usage As a guideline: the more context, files, data, and intermediate steps are processed, the more AI usage is consumed. **Normal usage**, covered through the Seat: * Chat * Knowledge search * Summaries * Meetings * Simple documents **Higher consumption,** plan additional capacity for regular use: * Large files * Complete presentations * Complex workflows * Building apps * Visualizations * API usage **Separate planning**, through dedicated Additional AI Usage: * External chatbots * Company-wide automations *** ## **AI Usage Tips** Powerful models consume more AI usage, and so do Apps, videos, and images. The greatest impact comes from deliberate use following one principle: think first, then prompt, then run. With Apps, videos, and images in particular, a short preparation step noticeably reduces consumption. * **Small context:** The biggest driver is usually not the prompt itself, but the conversation history carried along and any attached documents. For a new topic, a new chat works better than a long thread, since every message pulls in the entire prior context. For long files, the relevant excerpt is often enough instead of the whole document. * **Precise over vague prompts:** Imprecise requests lead to follow-up questions, wrong assumptions, and rework, and each round costs the full context again. A clear prompt that states context, goal, and desired format saves several clarification rounds. * **Output length control:** An explicit request for brevity ("keep it short", "bullet points only", "no explanation") reduces the output and often the follow-up rounds as well. * **No redundancy:** Large files or data sets do not need to be re-inserted with every message once they are already in the context. For recurring tasks, a template or a script works better than explaining the full context again. * **Bundled tasks:** Several small requests in sequence cost more in total than one structured request that covers everything, since each new message counts the previous history again. * **Only relevant tools and data:** Every unnecessarily loaded tool and every unnecessarily searched data source creates additional overhead. A targeted request is more economical than a broad search across all sources. * **Concept work first:** For slides or images, develop a script in a simple chat before starting generation. The finished script then serves as a precise prompt instruction and saves AI usage, because the AI has less room for interpretation and delivers the result on the first attempt. *** ## FAQ > **Which Plans and Seats are available?** > > Available are the Business Plan from €25 per user and month in self-serve, and the Enterprise Plan from 20 users with individual consultation. Within both Plans, the Seat-Types Standard (1x), Premium (2x), and Power (5x) are available. > **Are there feature-scope restrictions depending on Seat-Type or booked AI usage?** > > No. Seat-Types and Additional AI Usage differ only in the amount of included AI usage, not in feature scope. A higher Seat-Type or Additional AI Usage only expands the usage volume and does not unlock additional features. > **Do additional costs arise automatically with high usage?** > > No. Additional costs arise only when Additional AI Usage is deliberately booked or approved. > **Is nuwacom blocked when AI usage is exhausted?** > > No. Chat automatically switches to Auto Mode on a more cost-efficient model. Only external chatbots and workflows are blocked once their Additional AI Usage is exhausted. > **How is the AI usage of workflows billed?** > > A workflow always runs on behalf of a specific user and consumes that user's included or additionally booked AI usage, including automatic or recurring runs. For regular, complex, or company-wide automations, Additional AI Usage is recommended. > **Does uploading data to the Knowledge Base consume AI usage?** > > No. The manual upload consumes no AI usage. It arises only with AI actions on the data, such as search, summary, analysis, chat over knowledge, or app and workflow execution. > **Does the included AI usage renew monthly, and does unused usage carry over?** > > The included AI usage renews fully at the beginning of each month. Unused usage does not carry over. Additional AI Usage, in contrast, does not expire at the end of the month but remains usable throughout the contract term. > **Whose AI usage is consumed when an agent is shared?** > > Within the platform, AI usage is attributed to the person who **runs** the agent — not the person who created it. When an agent is shared with a team member and that team member sends a message, the usage is charged to their account.\ > \ > Note: For externally published agents — for example, embedded as a chat widget via iframe — usage is attributed to the agent owner. Every conversation initiated through the public endpoint counts against the owner's AI usage, regardless of who the end user is. # Identifying Use Cases Source: https://docs.nuwacom.ai/en/aiadoption/identifyingusecases This guide helps you recognize meaningful AI use cases in your company and implement them successfully. ### What makes a good AI use case? Strong use cases are situations where AI agents or workflows improve the quality of your work or product and/or reduce effort and time to results. **Tip:** Start with horizontal use cases—those relevant across teams. They offer several advantages: 1. **Shared understanding:** Acceptance increases when everyone can relate to the problem. 2. **Less customization:** Horizontal use cases typically require fewer individual adjustments. 3. **Faster start:** Highly specialized use cases demand more know-how. Begin with simple, scalable solutions to build experience. **Examples:** Email assistant, document summarization, translations—practical in almost every organization and significantly easing everyday work. *** ## How to find suitable use cases ### **1. Experiment & understand AI capabilities**  ![Frame 1254.png](https://mintcdn.com/nuwacomgmbh/y7RckkpWyYIF7ayw/images/Frame1254.png?fit=max\&auto=format\&n=y7RckkpWyYIF7ayw\&q=85\&s=7736554e1f0d2a74ef460b492ffede77) \ Let users experiment with AI and showcase concrete examples. New ideas for agents and workflows will emerge organically. **Typical AI capabilities:** | **Text** | **Images** | **Audio (coming soon)** | **Data Analysis** | | :- | :- | :- | :- | | Write | Create | Transcribe | Extract data | | Summarize | Analyze | Speak | Perform analyses and calculations | | Analyze | Describe | | Identify patterns | | Answer questions | Extract text | | Create tables and diagrams | ### **2. List daily task** Ask users to write down 5 recurring, time-consuming tasks from their workday. ### **3. Collect & cluster tasks** Gather tasks in a team session and group similar activities. Overlaps are opportunities for shared solutions. > [!NOTE] > > Tip: Use digital whiteboards (e.g., Miro, Mural, FigJam) for idea collection. ### **4. Match tasks with AI capabilities** Link the collected tasks to the right AI capabilities. Example: translations require text generation, summaries require text analysis and compression. ### **5. Prioritize: What to implement first?** ![Impact-Feasibility-Matrix.png](https://mintcdn.com/nuwacomgmbh/y7RckkpWyYIF7ayw/images/Impact-Feasibility-Matrix.png?fit=max\&auto=format\&n=y7RckkpWyYIF7ayw\&q=85\&s=ed23fb10dc0a77b8e58d90bfc865eb19) Don’t try everything at once. Focus on a few impactful use cases. A proven approach is using a 2x2 matrix with Feasibility and Impact: **Feasibility:** * **Effort:** Low effort = faster implementation * **Data availability:** Are the necessary data available and usable? * **Integrations:** Do interfaces or APIs need to be connected? **Impact:** * **Time savings:** How many hours can be saved? * **Quality improvement:** Will the result be better, more accurate? * **Customer satisfaction:** Will service quality improve? * **Financial effect:** Can costs be reduced or revenue increased? > [!NOTE] > > **Start with low-effort, high-impact use cases.** ### **6. Document & implement** Record results in a table listing all use cases, required AI capabilities, effort, impact, owners, and next steps. Here’s the extracted text from the table, translated into English: | ID | Opportunity / Problem | Solution Space | Use Case Title | Value | Ease | | :- | :- | :- | :- | :- | :- | | 01 | High time effort for FAQ maintenance | Knowledge Search & Chat | FAQ Auto-Updater | 4 | 3 | | 02 | Long response times in support | Text automation | Response Generator for Support Tickets | 5 | 4 | | 03 | Time-consuming market trend research | Research automation | AI-powered Trend Summary | 3 | 2 | ### **7. Build use cases together** Implement first cases as a team to build familiarity. Afterwards, let users work independently or in small groups. 1–2 weeks is ideal for first experiments. **Follow-up:** Support users, gather feedback, and share best practices in team meetings. > [!NOTE] > > **Important:** Not every use case works perfectly right away—this is part of the learning process. Stay open to adjustments. ## Summary With [nuwacom](https://nuwacom.com), you create the foundation for productive, secure, and scalable AI adoption by identifying use cases that deliver real value and can be implemented with ease. # Rollout Playbook Source: https://docs.nuwacom.ai/en/aiadoption/rolloutplaybook ## 1. Strategic preparation A successful AI rollout starts with clear goals and structured planning: * **Define objectives:** Decide which improvements you want to achieve with AI (e.g., more efficient processes, better collaboration, automation of routine tasks). * **Assign responsibilities:** Appoint a central AI Owner or team to coordinate the rollout and serve as the main contact. * **Governance & compliance:** Create internal guidelines, ensure data protection, and involve stakeholders (e.g., data protection officers) early. * **Technical foundation:** Review existing IT infrastructure, set up SSO, and integrate [nuwacom](https://nuwacom.com) with your current systems. ## 2. Onboarding & first use cases The start is crucial for acceptance and motivation: * **Kickoff session:** Begin with an interactive workshop, introduce the platform, and show relevant use cases. * **Prompt and assistant library:** Pre-fill the platform with practical examples for everyday work. * **Communication channels:** Set up a shared Teams/Slack channel for quick exchange and support. * **Pilot groups:** Select motivated users from different areas to act as multipliers. ## 3. Enablement & community building Building knowledge, fostering enthusiasm: * **Training & enablement:** Offer practical trainings, mini-webinars, and self-learning materials. Use a train-the-trainer approach so power users can share knowledge internally. * **Use case library:** Document and share successful use cases for cross-team learning. * **Regular deep dives:** Explore topics like prompt engineering, assistant building, and use case development in short, interactive sessions. * **Feedback culture:** Encourage open exchange, collect input, and continuously adapt content to user needs. ## 4. Scaling & long-term adoption From pilot to company-wide adoption: * **Onboard more users:** Gradually expand beyond the pilot phase to additional teams. * **Expand use cases:** Encourage teams to identify and implement their own cases (e.g., through workshops for idea generation and prioritization). * **Integrate more tools:** Add systems and data sources step by step to increase platform value. * **Community events:** Host hackathons, challenges, or AI days to maintain momentum and inspire innovation. ## 5. Measuring success & continuous improvement Transparency and iteration secure long-term success: * **Define KPIs:** Track active users, created use cases, time savings, or satisfaction. * **Make success visible:** Share best practices, celebrate milestones, and recognize engaged employees. * **Regular reviews:** Assess processes, adjust strategy, and scale proven approaches. * **Risk management:** Identify challenges early and develop contingency plans. ## 6. Avoiding common pitfalls * Don’t start with too many use cases at once—begin focused, test, then scale. * Avoid overregulation—keep processes lean and flexible. * Don’t neglect communication—regularly share progress, goals, and successes. ## 7. Practical measures for sustainable adoption * **Onboarding toolkit:** Provide checklists, guides, and sample workflows for a smooth start. * **Knowledge nuggets:** Share short tips and tricks regularly. * **AI newsletter:** Communicate updates, success stories, and new use cases company-wide. * **Reminders & motivation:** Provide small impulses in everyday work, e.g., digital notes or challenges. ## Conclusion Your roadmap for a successful AI rollout with [nuwacom](https://nuwacom.com): 1. Plan strategically and define responsibilities 2. Enable user-centered onboarding and early wins 3. Spread knowledge and enthusiasm across the organization 4. Scale, integrate, and continuously improve 5. Measure, share, and celebrate success With this playbook, you lay the foundation for productive, secure, and sustainable AI adoption in your company—tailored, practical, and future-oriented. *** Additional templates, checklists, and best practices can be found in the nuwacom-Help Center. Contact us if you’d like individual support to [support@nuwacom.ai](mailto:support@nuwacom.ai). # Rollout Setup Source: https://docs.nuwacom.ai/en/aiadoption/rolloutsetup ## Introduction: What makes an AI rollout unique? Introducing AI is fundamentally different from rolling out traditional software. While classic tools like CRM or ERP systems work deterministically (same input = same output), AI is probabilistic: the same request can lead to different results. This opens up enormous potential for customization, but it also requires targeted enablement and guidance for users. **Benefits:** 1. AI is flexible and can support almost every area of a business. 2. It fosters innovation, efficiency, and new forms of collaboration. **Challenges:** 1. Users need to learn how to work confidently with the new technology. 2. Clear responsibilities and a structured rollout process are essential. *** ## The ideal setup ### 1. Leadership Buy-in: Driving adoption from the top **Success factor:**\ AI integration works best when leadership actively drives the initiative. This means: 1. **Strategic anchoring:** AI is part of the overall business strategy, not just an IT project. 2. **Role modeling:** Leaders use AI themselves and communicate openly about opportunities and goals. 3. **Resources:** Time, budget, and staff are allocated for AI adoption. > [!TIP] > > **Best practice:**\ > Regular updates and success stories from the rollout create transparency and motivate the entire organization. ### 2. Internal AI responsibility: AI Owner & Team **Recommendation:**\ Appoint a central AI Owner or a dedicated AI team. This role is the hub for: 1. Coordinating the rollout 2. Identifying and prioritizing use cases 3. Training and supporting employees 4. Acting as the link between business units, IT, and management **Benefits:** 1. Quick response to questions or issues 2. Continuous development of AI usage across the company ### 3. AI Champions: Multipliers in business units **Why Champions?**\ Each department has different needs and AI potential. AI Champions are motivated employees who: 1. Test AI applications first and provide feedback 2. Support and encourage their colleagues 3. Act as a bridge between the AI Owner and their department **How to find Champions:** 1. Run an open call or directly approach AI power users 2. Provide incentives through training opportunities or recognition ### 4. User-centered adoption: Training & Enablement **Key points:** 1. Training should be practical and role-based (e.g., HR, Sales, Customer Service). 2. Hands-on workshops and use case sessions build understanding and acceptance. 3. A central Help Center provides accessible guides, FAQs, and best practices at any time. 4. AI chatbots can support onboarding for new employees while demonstrating AI’s potential. > [!TIP] > > **Best practice:**\ > Offer short “learning nuggets” and regular Q\&A sessions to lower barriers. ### 5. Use Case Development: From pilots to scaling **Approach:** 1. Start with well-defined pilot projects in selected departments. 2. Gather experience and measure success (e.g., time savings, quality improvements). 3. Scale successful approaches across the organization. **Best practice:**\ Document use cases and share them so all teams can benefit. ### 6. Governance & Data Protection: Security as a foundation **Essential:** 1. Clear guidelines for AI usage (e.g., handling sensitive data, result transparency). 2. Collaboration with data protection officers and IT security. 3. Regular reviews and updates of policies. ### 7. Continuous improvement: Feedback & Community **Recommendation:** 1. Establish feedback channels (e.g., via Help Center or surveys). 2. Promote an open learning culture and experience sharing (e.g., “My AI highlight of the week”). 3. Stay up to date with AI developments and adapt your strategy continuously. ## Summary Your roadmap for a successful AI rollout: 1. Secure leadership buy-in 2. Establish AI Owner and team 3. Recruit AI Champions in departments 4. Provide user training and enablement 5. Start pilots and scale successful use cases 6. Ensure governance and data protection 7. Gather feedback and continuously improve > **Tip** > > Use the nuwacom platform as your central operating system for productive, secure, and sovereign AI usage. This creates the foundation for sustainable success and a future-ready workplace. *** Further resources, templates, and best practices can be found in the nuwacom-Help Center. We’re happy to support you in planning and executing your individual AI rollout. ![nuwacom\_solution.png](https://mintcdn.com/nuwacomgmbh/uiYWlDKEm7UfCFMt/images/nuwacom_solution.png?fit=max\&auto=format\&n=uiYWlDKEm7UfCFMt\&q=85\&s=7038845438f3758904a597e6f3712576) # AI Models Source: https://docs.nuwacom.ai/en/aimodels The latest and most advanced AI models—seamlessly integrated ## **What are AI models?** ![AI models.jpg](https://mintcdn.com/nuwacomgmbh/QmokFyROSwNjqeyr/images/AImodels.jpg?fit=max\&auto=format\&n=QmokFyROSwNjqeyr\&q=85\&s=b99ca2979dc5ded93d867f30ddf44b4f) AI language models are intelligent systems that can understand, generate, and analyze text. They help individuals and businesses efficiently create content, process complex information, and make informed decisions. Especially in marketing and corporate communications, they prove extremely useful by optimizing texts, speeding up research processes, and preparing versatile content. In nuwacom, you can choose the model that best suits your specific requirements. ## **Which AI models does nuwacom offer?** ### Auto Mode Auto Mode is enabled by default in chat. nuwacom automatically selects the model that best fits each request: the Instant model for simple tasks and a more powerful model for more complex requests. The currently assigned models are shown under *Workspace Settings → Models* and are regularly updated by nuwacom based on performance, speed, and efficiency. Select another model at any time via the dropdown menu if needed. nuwacom provides access to a wide range of modern AI models, flexible to use, securely hosted, and tailored to real-world use cases. The following overview shows the available model providers and models in nuwacom. ### Models | **Model** | **Provider** | **Context Window** | **EU Hosting** | **Ideal for** | | :- | :- | :- | :- | :- | | GPT-6 Astra | OpenAI | 1M | ✅ 🇪🇺 | OpenAI's most capable model, with advanced reasoning and cybersecurity capabilities | | GPT-6.1 Sol | OpenAI | 1M | ✅ 🇪🇺 | Near-Astra intelligence for agentic work, at a fraction of the cost | | GPT-6 Sol | OpenAI | 1M | ✅ 🇪🇺 | Balanced GPT-6 for reasoning and speed | | GPT-6 Luna | OpenAI | 1M | ✅ 🇪🇺 | Fast, cost-efficient GPT-6 | | GPT-5.6 Sol | OpenAI | 1M | ✅ 🇪🇺 | Most capable GPT-5.6 for complex reasoning | | GPT-5.6 Terra | OpenAI | 1M | ✅ 🇪🇺 | Balanced GPT-5.6 for reasoning and speed | | GPT-5.6 Luna | OpenAI | 1M | ✅ 🇪🇺 | Fast, cost-efficient GPT-5.6 | | GPT-5.5 | OpenAI | 1M | ✅ 🇪🇺 | Latest flagship model for complex coding and professional tasks | | GPT-5.4 | OpenAI | 1M | ✅ 🇪🇺 | Flagship model for professional and enterprise use | | GPT-5.2 | OpenAI | 400k | ❌ 🇺🇸 | Coding and agentic tasks across different industries | | GPT-5.1 Instant | OpenAI | 128k | ❌ 🇺🇸 | Fast and cost-efficient reasoning | | GPT-5.1 | OpenAI | 400k | ✅ 🇪🇺 | Advanced reasoning | | GPT-5 | OpenAI | 400k | ✅ 🇪🇺 | Best model for coding and agentic tasks across different domains | | GPT-5 mini | OpenAI | 400k | ✅ 🇪🇺 | A faster and more cost-efficient version of GPT-5 for clearly defined tasks | | GPT-5 Nano | OpenAI | 400k | ❌ 🇺🇸 | Fastest and most cost-efficient version of GPT-5 | | GPT-4.1 | OpenAI | 1M | ✅ 🇪🇺 | Ideal for complex tasks | | GPT-4.1 Mini | OpenAI | 1M | ✅ 🇪🇺 | Fast for everyday tasks | | GPT-4.1 Nano | OpenAI | 1M | ❌ 🇺🇸 | Fastest and most cost-efficient GPT-4.1 model | | GPT-4o | OpenAI | 128k | ✅ 🇪🇺 | Ideal for most tasks | | GPT-4o mini | OpenAI | 128k | ✅ 🇪🇺 | Fast for everyday tasks | | o3-mini | OpenAI | 200k | ✅ 🇪🇺 | Faster, improved reasoning | | o4-mini | OpenAI | 200k | ✅ 🇪🇺 | Effective, fast reasoning | | o1 | OpenAI | 200k | ✅ 🇪🇺 | Advanced reasoning | | o3 | OpenAI | 200k | ✅ 🇪🇺 | Advanced reasoning | | Claude Fable 5.1 | Anthropic | 1M | ✅ 🇪🇺 | Anthropic's most capable model | | Claude Fable 5 | Anthropic | 1M | ✅ 🇪🇺 | Latest and most skilled model from Anthropic | | Claude Opus 5.5 | Anthropic | 1M | ✅ 🇪🇺 | Anthropic's latest Opus for agents and coding, at a lower price | | Claude Opus 5 | Anthropic | 1M | ✅ 🇪🇺 | Anthropic's most advanced Opus for long-running agents and coding | | Claude Opus 4.8 | Anthropic | 1M | ✅ 🇪🇺 | Anthropic’s latest and most powerful model | | Claude Opus 4.7 | Anthropic | 1M | ✅ 🇪🇺 | Advanced reasoning and agentic capabilities | | Claude Opus 4.6 | Anthropic | 1M | ✅ 🇪🇺 | Advanced reasoning and agentic capabilities | | Claude Opus 4.5 | Anthropic | 200k | ✅ 🇪🇺 | Intelligent, efficient, and better for everyday tasks | | Claude Sonnet 5.5 | Anthropic | 1M | ✅ 🇪🇺 | Anthropic's latest Sonnet, fast and capable for coding and documents | | Claude Sonnet 5 | Anthropic | 1M | ✅ 🇪🇺 | Near-frontier intelligence for coding and agentic work | | Claude Sonnet 4.6 | Anthropic | 1M | ✅ 🇪🇺 | Best for computer use, planning, and knowledge work | | Claude Sonnet 4.5 | Anthropic | 200k | ✅ 🇪🇺 | Anthropic’s latest flagship model | | Claude Haiku 5.5 | Anthropic | 1M | ✅ 🇪🇺 | Anthropic's fastest and most efficient model for high-volume tasks | | Claude 4.5 Haiku | Anthropic | 200k | ✅ 🇪🇺 | Fastest near-frontier intelligence | | Gemini 3.8 Flash | Google | 1M | ✅ 🇪🇺 | Google's latest Flash model for coding and agents | | Gemini 3.7 Flash | Google | 1M | ✅ 🇪🇺 | Most intelligent Flash model for coding and agents | | Gemini 3.6 Flash | Google | 1M | ✅ 🇪🇺 | Optimized for agentic tasks, code generation, and spatial reasoning | | Gemini 3.5 Flash | Google | Google | ✅ 🇪🇺 | Near-Pro intelligence at Flash speed and cost | | Gemini 3.5 Flash-Lite | Google | 1M | ✅ 🇪🇺 | Fastest, most cost-effective Gemini for high-volume tasks | | Gemini 3 Pro Preview | Google | 1M | ❌ 🌎 | Everyday tasks | | Gemini 2.5 Pro | Google | 1M | ✅ 🇪🇺 | Most advanced Gemini model for logical reasoning | | Gemini 2.5 Flash | Google | 1M | ✅ 🇪🇺 | Fast and efficient Gemini model with thinking capabilities | | Gemini 2.5 Pro Reasoning | Google | 1M | ✅ 🇪🇺 | Gemini model for deep logical reasoning | | Mistral Large 3 | Mistral | 128k | ✅ 🇪🇺 | Ideal for most tasks | | GLM 5.3 | Z.ai | 1M | ✅ 🇪🇺 | Sovereign EU-hosted reasoning model | | GLM 5.3 Flash | Z.ai | 1M | ✅ 🇪🇺 | Fast sovereign EU-hosted model with image and video input | | GLM 5.2 | Z.ai | 1M | ✅ 🇪🇺 | Sovereign EU-hosted reasoning model | | GLM 5.1 | Z.ai | 203k | ✅ 🇪🇺 | Sovereign EU-hosted reasoning model (previous version) | | GLM 5 | Z.ai | 203k | ✅ 🇪🇺 | Sovereign EU-hosted reasoning model (earlier version) | | DeepSeek V4 Flash 0731 | DeepSeek | 1M | ✅ 🇪🇺 | Sovereign EU-hosted reasoning model | | DeepSeek V4 Pro 0813 | DeepSeek | 1M | ✅ 🇪🇺 | Sovereign EU-hosted reasoning model | | Kimi K3 | Moonshot | 1M | ✅ 🇪🇺 | Sovereign EU-hosted reasoning model | | DeepSeek V4.1 Flash | DeepSeek | 1M | ✅ 🇪🇺 | Sovereign EU-hosted model with image input | | DeepSeek V4.1 Flash 1017 | DeepSeek | 1M | ✅ 🇪🇺 | Sovereign EU-hosted reasoning model with image input | | Qwen 3.8 Max | Qwen | 262k | ✅ 🇪🇺 | Most capable sovereign EU-hosted Qwen reasoning model | | Qwen 3.8 27B | Qwen | 262k | ✅ 🇪🇺 | Fast sovereign EU-hosted reasoning model with image input | | Qwen 3.6 27B | Qwen | 262k | ✅ 🇪🇺 | Compact sovereign EU-hosted reasoning model with image input | | MiniMax M3 | MiniMax | 1M | ✅ 🇪🇺 | Sovereign EU-hosted agentic model with image input | ### **Image generation models** * Gemini 2.5 Flash Image (Nano Banana) (EU-hosted) * Gemini 3.1 Flash Image (Nano Banana 2) (EU-hosted) * GPT Image 2.0 (EU-hosted) * GPT Image 2.5 (EU-hosted) * FLUX 1.1 Pro Ultra (EU-hosted) * FLUX 1 Kontext Pro (EU-hosted) * FLUX 2 Pro (EU-hosted) * Gemini 3 Pro Image (Nano Banana Pro) (Global-hosted) * Gemini 3.1 Flash Image (Nano Banana 2) (Global-hosted) > [!NOTE] > > Gemini 3.1 Flash Image is (Nano Banana 2, Global-hosted) used by default. As an admin, you can select another image model in Workspace Settings > Models. ### **Video generation models** * **Veo 2** (EU-hosted) * **Veo 3** (US-hosted) > [!NOTE] > > Veo 2 (EU) is used by default for video generation in your workspace. As an admin, you can change the model if needed in Workspace Settings > Models > Video generation model. ## **What are context windows?** A context window defines how much text an AI model can “see” and process at one time—depending on the model, for example: * 8,000 tokens ≈ 4 pages * 32,000 tokens ≈ 16 pages * 128,000 tokens ≈ 64 pages * 1,000,000 tokens ≈ 600 pages Even large context windows have their limits with very long documents. > [!NOTE] > > **At nuwacom, we overcome this limitation with RAG (Retrieval-Augmented Generation):** > > the AI intelligently searches through large datasets and extracts only the relevant sections, no matter how long the document is. > > **Example:** You ask, “What data protection measures are specified in the contract?” Instead of analyzing all 2,000 pages across 10 documents, the AI jumps directly to the relevant paragraphs. You receive a precise, clear answer in seconds. ## **What are the benefits of model selection?** The key benefit of flexible model selection in nuwacom is that you can choose the model best suited for each specific task. This makes it possible to handle creative writing, highly specialized technical texts, and routine tasks efficiently. This versatility ensures precise, effective use of AI, improves result quality, and enhances collaboration among project teams with diverse requirements. ## **Which model is suitable for which tasks?** ![Chattasksen](https://mintcdn.com/nuwacomgmbh/osV_aASm-7aDJPwy/images/Chattasksen.avif?fit=max\&auto=format\&n=osV_aASm-7aDJPwy\&q=85\&s=9130ccd2f63275d15aa193b2c9ef29df) Each AI model has its own strengths. With nuwacom, you do not need to decide which model is best suited for your task yourself. The platform is designed to be LLM-agnostic from the ground up and automatically selects the model that best fits your use case. Simply select the right task to structure your request and get concrete results, for example: * **Use the Orchestrator** – Start complex tasks that are automatically planned, coordinated, and handled by specialized agents. * **Create slides** – Generate presentation slides for meetings or pitches. * **Create a document** – Write structured texts such as concepts or reports. * **Build workflows** – Automate recurring processes by connecting multiple steps. * **Write a note** – Capture ideas or meeting notes. * **Research** – Conduct structured research on markets, competitors, or trends. * **Search company knowledge** – Search company knowledge and find relevant information from connected sources. * **Generate images** – Create visual content from a description. * **Generate videos** – Create videos based on text descriptions. * **Analyze data** – Analyze or summarize datasets. * **Create an agent** – Create custom AI agents that support recurring tasks. * **Draft an email** – Write professional emails. > [!TIP] > > Depending on the selected use case, the optimal AI model is selected automatically. This helps users without specific prior knowledge get precise results efficiently and in line with each task. *** ## **Hosting** ### **🇪🇺 EU-Hosted Models** These models are operated exclusively on servers within the European Union. **What are the advantages of EU-hosted models?** * **GDPR compliance:** Meets the requirements of the General Data Protection Regulation. * **No transfer to third countries:** Your input and generated content never leave the EU. * **Suitable for sensitive data:** Ideal for companies working with personal or confidential information. **When is this the right choice?**\ If you use language models for internal communication, customer support, or data analysis—and need to ensure no data is transferred to the USA. ### **🇺🇸 US-Hosted Models** These models run on servers located in the United States. **What are the advantages of US-hosted models?** * **Often more up-to-date or powerful:** New or experimental models are usually available first in the USA. * **Broader model selection:** Access to advanced features. **What should you consider?**\ Usage may involve higher data protection risks, such as access by US authorities (Cloud Act). For some organizations, this can present a compliance risk. ### **Which hosting option is right for me?** * Choose an **EU-hosted** model if data protection, GDPR compliance, and regulatory security are particularly important to you. * Choose a **US-hosted** model if you want access to the latest features and models—and you are not working with sensitive data. *** ## **Fair Use Policy** [Open](https://a.storyblok.com/f/290757788231937/x/6988009b94/fairusepolicy_nuwacom_290825_english-1.pdf) # Analytics Source: https://docs.nuwacom.ai/en/analytics ## Overview ![Analyticsen](https://mintcdn.com/nuwacomgmbh/n4lKCOY9l7_aCUKc/images/analyticsen.png?fit=max\&auto=format\&n=n4lKCOY9l7_aCUKc\&q=85\&s=cce30460f7344183bdc6af7b4bba36c1) Analytics provides workspace administrators with aggregated insights into usage across Chat, Agents, Workflows, and the nuwacom API. Use it to monitor adoption, compare activity over time, and export chart data for reporting. Analytics are available in the workspace settings. ## Key benefits * Track workspace usage across chats, Agents, Workflows, and the nuwacom API. * Review central workspace metrics, including users, Agents, Workflows, and documents. * Compare activity over time by grouping data by day, week, or month. * Export chart data as CSV for analysis in Excel, reporting tools, or internal reports. ## When to use Analytics Analytics help admins understand how a workspace is used over time. The area supports usage reviews, internal reporting, adoption tracking, and function-level analysis across Chat, Agents, Workflows, and API usage. *** ## How to view analytics data 1. Open Workspace settings. 2. Select Analytics. 3. Choose an analytics tab. 4. Set the Date range and Group by filters. *** ## Setting the date range and grouping The filters above the analytics view control how data appears across charts and metrics. When a filter changes, all affected charts and metrics update based on the selected settings. ### Group by **Group by** defines the time interval used to aggregate the data. Available examples include day, week, and month. ### Date range **Date range** defines the period used for the usage statistics. Available examples include today, last week, and a custom range. *** ## Analytics areas The Analytics area contains separate tabs for different evaluation areas. The available tabs are **Overview**, **Chat**, **Agents**, **Workflows**, and **API**. ### Overview The Overview tab shows central workspace metrics. These include Users in workspace, Total Agents, Total Workflows, and Total documents. | Metric | Description | | :- | :- | | **Active users** | Shows the number of individual users who sent at least one message | | **All messages** | Shows all user messages across all conversations | Moving the cursor over a chart bar displays the exact value for the selected point in time. ### Chat The Chat tab shows how Chat was used within the selected date range. The data follows the selected grouping, for example by day. | Metric | Description | | :- | :- | | **All messages** | Shows the total number of user messages across all conversations. | | **Model usage distribution** | Shows which AI models were used for generated answers. | | **Knowledge Base searches** | Shows the number of AI answers that retrieved content from connected knowledge bases. | Message-based metrics count user messages. Generated AI answers are reflected separately in model usage metrics. ### Agents The Agents tab shows how Agents were used within the selected date range. Evaluations follow the selected date range and grouping, for example by day. | Metric | Description | | :- | :- | | **Agent messages** | Shows the number of user messages sent in conversations with an Agent. | | **Active Agent users** | Shows the number of individual users who sent at least one message in an Agent conversation. | | **Messages per Agent** | Shows how many user messages were sent to each Agent. | | **Agent model usage** | Shows which AI models were used for answers in Agent conversations. | Agent filters are available for the *Active agent users* and *Messages per agent* metrics. Selecting one or more agents makes it possible to analyze how a specific combination of agents was used. ### Workflows The **Workflows** tab shows how Workflows were executed within the selected date range and which status each execution ended with. The data follows the selected date range and grouping, for example by day. The status overview summarizes the main execution statuses above the chart. | Metric | Description | | :- | :- | | **Completed executions** | Shows successfully completed Workflow executions. | | **Failed executions** | Shows Workflow executions that did not complete due to an error. | | **Cancelled executions** | Shows Workflow executions that ended early. | | **Workflow executions by status** | Shows the number of Workflow executions over time. The colored bar sections show how statuses are distributed across each time interval. This makes periods with many completed, failed, or cancelled executions easier to identify. | ### API The **API** tab shows usage of the nuwacom API within the selected date range. The evaluation follows the selected date range and grouping, for example by day. | Metric | Description | | :- | :- | | **Total API calls** | Shows how many API requests were executed in total within the selected date range. This provides a compact view of overall API usage in the workspace. | | **API calls over time** | Shows the number of API calls for each time interval. This makes visible when the API was used most often and how usage developed within the selected date range. | *** ## Exporting data Select **Export CSV** on a chart to download the data currently represented by that chart. Each export contains one row per data point (for example, one row per model) and the corresponding aggregated count for the selected date range. The filename includes the chart type and the selected time period. The exported data can be processed outside nuwacom, for example in Excel, a reporting tool, or internal analyses. *** ## Additional metric information The information icon next to an evaluation opens a short description of the corresponding metric. This helps verify what a chart or metric represents before the data is used in reports. *** ## FAQ > **Who can access Analytics?** > > The Analytics area is visible only to workspace admins. Open **Workspace settings**, then select **Analytics**. > **Which areas can be evaluated in Analytics?** > > Analytics includes **Overview**, **Chat**, **Agents**, **Workflows**, and **API**. These areas provide general workspace metrics as well as insights into individual functions. > **Which API usage is displayed?** > > The **API** tab shows the total number of API calls and their development over the selected date range. This shows when the API was used and how usage changed. > **How can the date range and grouping be changed?** > > Use the filters above the analytics view. Select a date range such as today, last week, or a custom range. Group the data by day, week, or month. > **Can Analytics data be exported?** > > Data from each chart can be exported individually as a CSV file. The export reflects the selected date range and grouping. The file can then be used in Excel, reporting tools, or internal analyses. > **Can admins view usage data for individual users based on their email addresses?** > > No. For data protection reasons, nuwacom does not provide usage data at the level of individual users or email addresses. All metrics are displayed in anonymized and aggregated form only. > > To assess individual license usage, teams can use internal discussions or surveys as an additional source of information. > **Does Analytics include usage from embedded chats or widgets?** > > No. Analytics currently only includes usage within the nuwacom platform. Requests made through embedded chats or widgets via iFrame are not included in the Analytics data. # Connect an MCP Server Source: https://docs.nuwacom.ai/en/connect-mcp-server ## Overview MCP servers enable deeper and more flexible integrations beyond standard Apps. MCP stands for Model Context Protocol and provides a standardized way to connect external systems. Agents can then access data and run more complex, granular actions across systems. MCP servers suit advanced use cases, custom integrations, and technical workflows. A curated list of common MCP servers is available, and custom servers can be added manually when needed. ## Connect an MCP server Any MCP server can be connected from the workspace settings, whether it comes from the catalog or is a custom server you host yourself. A short wizard inspects the server, handles authentication, and enables selection of the exact tools available to agents. ### Step by step 1. Open the workspace settings, select *MCP Servers*, and choose *Add server*. 2. Enter the server URL and, optionally, a display name, then select *Verify*. nuwacom inspects the server and detects its transport type (StreamableHTTP or SSE), whether authentication is required, and which authentication method applies. 3. Name the connection and provide the credentials for the detected authentication method (see Authentication methods). 4. Select *Test connection*. nuwacom connects to the server and discovers the available tools. If the connection fails, the error returned by the server is displayed so the settings can be adjusted. 5. Select the tools to enable for agents, then confirm to create the connection. > [!NOTE] > > Enabled tools, the connection name, sharing, and removal can be changed at any time under Connectors in the workspace settings. ## Authentication methods nuwacom auto-detects how a server expects authentication and displays only the relevant fields: * No authentication: Public servers that require no credentials. The URL is verified and setup continues. * API key / token: The server is protected by a static API key or bearer token. A stored credential is selected for the connection. For OAuth servers, this method is also available via *Use API key / token instead* in the advanced options. * OAuth: The server uses OAuth 2.0. nuwacom suggests the best mode based on the server's discovery metadata (see the OAuth modes below). ## OAuth modes When a server uses OAuth, the method for obtaining a token can be selected: * **Automatic (recommended):** Sign-in runs through the browser. The client registers itself automatically with the provider via Dynamic Client Registration (DCR), so no credentials are entered. This works for most servers. * **Browser login with your own credentials:** Sign-in runs through the browser using a Client ID and Secret from an OAuth app registered with the provider. Use this when the server does not support automatic registration. * **Client Credentials (M2M):** The connection uses a Client ID and Secret directly, without browser login. This covers machine-to-machine (M2M) access such as CI/CD pipelines, server-to-server integrations, and automated workflows. It is available only for servers that support the OAuth client credentials grant. * **Manual configuration:** The Authorization URL, Token URL, Client ID, and Secret are entered manually. Use this when a server does not expose its OAuth discovery metadata. > **Tip:** To limit access, open *Advanced settings* in the connection step and specify custom OAuth scopes (space- or comma-separated). Leaving the field empty requests the scopes the server advertises. > [!NOTE] > > Not every server supports every mode. Many hosted MCP servers offer only browser-based login. Machine-to-machine (M2M) access requires the server to support the OAuth client credentials grant. # Connections Source: https://docs.nuwacom.ai/en/connections ## Overview A connection is a saved link between nuwacom and an external tool. Connecting a tool such as Google Calendar, HubSpot, or Slack creates a connection that stores the access needed to work with that service. Each connection belongs to the person who created it. By default it is private: only the owner can see and use it, and actions run with the owner’s own access rights in the external service. ## Authentication Types How you connect depends on the Connector. | Auth Type | How It Works | | :- | :- | | **OAuth** | Sign in with the service and grant nuwacom permission. | | **API key / token** | Paste an API key or token from the service into nuwacom. | | **None** | No authentication needed. | ### OAuth Most popular Connectors use OAuth, the industry-standard protocol for secure authorization: 1. nuwacom sends you to the service’s login page, for example Google or Microsoft. 2. Sign in with your own account. 3. Approve the requested permissions. 4. nuwacom stores the access securely. The connection acts on your behalf. If you cannot open a calendar in Google, an agent using your connection cannot open it either. ### API key / token Some Connectors, especially custom ones, are set up with an API key or token. Copy it from the service’s settings and paste it into nuwacom when connecting. *** ## Reconnecting If a connection has expired or been revoked, its account shows a warning icon with the note *Connection issue*. Reconnect the account to restore access. *** ## Who Can Use a Connection ### Private by default A new connection is shown as **Private account**: 1. Only the owner can see and use it. 2. Actions run with the owner’s access rights. 3. Credentials are never shown to other users. ### Sharing an account The owner can share an account from *Settings → Connectors*: 1. Find the account in the list. 2. Select the share icon next to it (*Share \{account name}*). 3. Choose who gets access: specific users, user groups, or all workspace users. Shared accounts are labeled **Shared account**. Sharing with the entire workspace is available if your role allows sharing with the workspace. Share an account only with people you trust to act through it. Everyone with access works with the owner’s permissions in the external service. *** ## Using a Default Account in an Agent You can let other people run a Connector action with a specific account without sharing the account itself. They can use it only through the agent. They never see its credentials. ### Configure it 1. Open *Agents* in the sidebar and open or create an agent. 2. In *Actions*, select *Add action* and choose the Connector action. 3. Select *Select default account*. 4. Choose the account and select *Add action*. 5. Publish the agent and share it with the people who should use it. ### The two modes 1. **Select default account**: One account is used for everyone who runs the action. Whatever the user’s own accounts are, the action runs with this account’s access. 2. **Every user must use their own account (advanced)**: Each person uses their own account. People who have not connected yet are asked to connect. ### Examples * **Default account:** A “Team Calendar Agent” uses a shared team calendar. When any colleague uses it, the event lands in the team calendar. * **Own account:** A “Personal Email Agent” sends mail from each user’s own mailbox, using their own connection. If no account is selected for an action, every user has to connect their own account when using the agent. *** ## Apps Apps always run with the connected accounts of the person who opens them. Your own connected accounts stay private and are never used by people you share an app with. *** ## Choosing the Right Approach | Scenario | Recommended Approach | | :- | :- | | Team needs a shared calendar or CRM account | Add the action to an agent with a default account | | Everyone should use their own mailbox or calendar | Agent with *Every user must use their own account* | | Colleagues should see and use the account themselves | Share the account from *Settings → Connectors* | | Personal workflow | Use your own connection in your own agent | | | | ## FAQ > **Who owns a connection?** > > The person who created the connection owns it. The connection remains private until it is shared with other users or the workspace. > **Why can’t a colleague use my connection in their agent?** > > Check whether the connected account is shared with the colleague or whether the agent has a default account selected. > > If neither applies, the colleague needs to connect their own account before they can use the Connector. > **Why does a Connector ask me to reconnect?** > > The connection may have expired or been revoked. > > Reconnect the account under *Settings → Connectors* to restore access. # Connector Directory Source: https://docs.nuwacom.ai/en/connector-directory The Connector Directory lists all connectors available in nuwacom, each accessible across Chat, Agents, Workflows, Knowledge, and App once connected. ## Available Connectors nuwacom includes a set of built-in connectors for the tools teams already use, covering communication, email, file storage, project management, knowledge bases, and CRM. Each handles authentication, data syncing, and API connections automatically, keeping the focus on AI usage rather than technical configuration. Connection runs via OAuth using an existing account: selecting a connector in the directory and providing the required authentication details completes the setup and makes it available everywhere. > **Microsoft Teams** > > **Available Actions:** Search messages across channels, review meeting transcripts, and post to conversations > > **Authentication Method:** OAuth – Sign-in via existing account > **Microsoft Outlook** > > **Available Actions:** Search, draft, organize, and send emails, plus review the calendar, find open slots, and create or reschedule events > > **Authentication Method:** OAuth Sign-in via existing account > **Microsoft Excel** > > **Available Actions:** Read data, manage tables, update cells, insert formulas, and build charts > > **Authentication Method:** OAuth Sign-in via existing account > **OneDrive** > > **Available Actions:** Search, browse, and download files, and upload new ones > > **Authentication Method:** OAuth Sign-in via existing account > **SharePoint** > > **Available Actions:** Search documents, read page content, manage lists, and upload files > > **Authentication Method:** OAuth Sign-in via existing account > **Power BI** > > **Available Actions:** Query datasets, explore reports, and export dashboards > > **Authentication Method:** OAuth Sign-in via existing account > **OneNote** > > **Available Actions:** Browse sections, search pages, and create new pages > > **Authentication Method:** OAuth Sign-in via existing account > **Microsoft Planner** > > **Available Actions:** Browse plans, track task statuses, and add or update tasks > > **Authentication Method:** OAuth Sign-in via existing account > **Gmail** > > **Available Actions:** Search messages, read attachments, organize labels, and compose or reply to emails > > **Authentication Method:** OAuth Sign-in via existing account > **Google Calendar** > > **Available Actions:** Review the schedule, search events, verify availability, and create or reschedule meetings > > **Authentication Method:** OAuth Sign-in via existing account > **Google Drive** > > **Available Actions:** Search files, browse folders, download files, and upload new ones > > **Authentication Method:** OAuth Sign-in via existing account > **Google Docs** > > **Available Actions:** Search documents, read content, and create or update docs > > **Authentication Method:** OAuth Sign-in via existing account > **Google Sheets** > > **Available Actions:** Read spreadsheet data, search across sheets, and create or update rows > > **Authentication Method:** OAuth Sign-in via existing account > **Google Meet** > > **Available Actions:** Retrieve meeting details and transcriptions > > **Authentication Method:** OAuth Sign-in via existing account > **Slack** > > **Available Actions:** Search messages, browse channels, look up the team directory, and post replies > > **Authentication Method:** OAuth Sign-in via existing account > **Jira** > > **Available Actions:** Search issues, verify statuses, read comments, and create or update tickets > > **Authentication Method:** OAuth Sign-in via existing account > **Confluence** > > **Available Actions:** Search content, read pages and comments, and create or update pages > > **Authentication Method:** OAuth Sign-in via existing account > **Notion** > > **Available Actions:** Search pages, query databases, read content, and create or update pages > > **Authentication Method:** OAuth Sign-in via existing account > **Pipedrive** > > **Available Actions:** Find deals, leads, people, and organizations, review activities and notes, and create or update entries > > **Authentication Method:** OAuth Sign-in via existing account > **HubSpot** > > **Available Actions:** Search and manage contacts, companies, deals, and tickets, review CRM activities, and create or update records > > **Authentication Method:** OAuth Sign-in via existing account > **Personio** > > **Available Actions:** Access and manage employee information, retrieve absence and attendance data, and manage HR records > > **Authentication Method:** OAuth Sign-in via existing account > **Google Analytics (GA4)** > > **Available Actions:** Access analytics properties, retrieve traffic reports, analyze pages, audiences, sources and conversions, monitor real-time activity, and compare reporting periods > > **Authentication Method:** OAuth Sign-in via existing account > **Monday** > > **Available Actions:** Search and manage boards, items, groups, and workspaces, analyze project data, track changes, and create or update tasks and comments > > **Authentication Method:** OAuth Sign-in via existing account # Creating Design systems Source: https://docs.nuwacom.ai/en/creating-design-systems Creating a Design system opens a chat beside its page. The page fills with the elements gathered during the conversation. Each creation chat builds one Design system. A website, brand files or a description in chat can provide the starting point. 1. Open *Design systems*. 2. Create a new Design system. 3. Select *Your website* or *Brand files* as the starting point. 4. Enter the website address or provide a presentation, template, PDF or Word file. 5. Review the overview, colors, typography, logos, graphics and files in the editor. A Design system can also start without a website or file. In that case, the builder asks for basic information about the brand in chat. Logos, graphics and font files can be added separately. ## Reviewing Design System After nuwacom processes the supplied sources, the Design system is organized into six tabs: * *Overview*: A summary of the Design system and its design guidance. * *Colors*: The captured color palette. * *Typography*: The captured text styles. * *Logos*: Brand logos. * *Graphics*: Other visual assets. * *Files*: Files added as sources. To check the result, select the Design system in chat and create a representative output, such as a landing page or presentation. Compare its appearance with the brand materials. ## Call a Design System A Design system can be selected on the chat start page or through *@ > Design systems*. If exactly one Design system is available in the workspace, nuwacom applies it automatically. If several are available and none has been selected, nuwacom asks which one to use. It does not combine two Design systems. The toggle on an individual Design system enables or disables its automatic use in a person’s own chats. A disabled Design system remains available for manual selection. ## Supported sources and limits The following file formats and sources are supported when creating a Design system: **Documents** * **PowerPoint:** .pptx, .potx * **Word:** .docx, .dotx * **PDF:** .pdf (first 30 pages) * **File size limit:** 50 MB per file **Websites** * Must be publicly accessible via HTTPS * Maximum review time: 45 seconds per website **Images** * .svg, .png, .jpg, .jpeg, .webp, .gif **Fonts** * .woff, .woff2, .ttf, .otf **Design System Limits** * **Colors:** Up to 12 * **Type styles:** Up to 24 * **Sources:** Up to 50 *** ## FAQ **What happens if a website cannot be read?** Only public HTTPS pages can be reviewed. If a site blocks automated access or takes too long to load, a supported brand file can be used instead. An unsuccessful review does not change the Design system. **Can a Design System be created without brand files or a website?** Yes. A description in chat provides the starting point. The builder asks for basic brand details and builds the Design system from the answers. Its elements can then be refined on the page or in chat. **Which fonts are retrieved automatically?** Freely available, license-free fonts, including suitable web fonts, are retrieved automatically. **Which fonts must be uploaded?** Fonts that require a license must be uploaded separately as font files. **What does the label *Font is missing* mean, and how is it resolved?** A font defined in the design system is not available to the system. This applies to licensed fonts without an uploaded font file and to fonts that cannot be retrieved automatically. To resolve it: 1. Open *Design systems* and select the affected design system. 2. Open the *Typography* tab. 3. Upload the font file for the font marked *Font is missing*. Once the file is uploaded, the font is available for documents, slides and apps created with this design system. *** # Creating Skills Source: https://docs.nuwacom.ai/en/creating-skills ## Creating a New Skill Skills are created from the sidebar. Opening Skills and selecting + New Skill opens a window in which the new Skill is created and configured. One of the following options applies. ## Build with AI (recommended) A plain-language description defines the Skill: what it should do, what tone to enforce, and what structure to apply. The input field also supports voice input. Once the description is sent, nuwacom generates a complete Skill with a name, description, and instructions. The generated Skill appears directly in the modal, where it can be previewed, inspected, and refined before saving. 1. Preview: The Preview button shows the Skill as it will appear when applied, including name, description, and formatted instructions. 2. Edit: The Edit option makes the generated name, description, and instructions adjustable before saving. 3. Try again: nuwacom generates a new version of the Skill instructions. 4. Save skill: The Skill is saved to the library, private by default, and editable at any time. ## Start from a Template The Skills feature also offers a template library. Templates can be browsed by category (HR, Legal, Product, Sales, Support, Writing) or found through the search field. A selected template becomes the starting point for the Skill. ## More Options The three-dot menu in the modal header provides additional creation paths. ### Build Skill from Scratch Name, description, and instructions are entered manually. This path suits well-defined use cases where the structure is already known. ### Import Skill An existing Skill file in a supported format (SKILL.md, .zip, or .skill) can be imported to share or reuse Skills across workspaces. ### Import from GitHub A pasted repository link imports a Skill directly. This suits teams working with version control. *** ## Add Your Own Files to Skills Skills support file attachments, so custom materials feed directly into content generation. Brand guidelines, logos, reference graphics, and fonts can be attached to a Skill and are applied automatically whenever the Skill generates images, documents, or presentations. ## Supported File Formats ### Text Files (editable in the Skill Editor) 1. Markdown (.md) 2. Python (.py) 3. Text (.txt) 4. JSON (.json) 5. YAML (.yaml, .yml) 6. CSV (.csv) 7. HTML (.html) 8. SVG (.svg) ### Asset Files (uploadable as attachments) 1. Images: PNG (.png), JPG/JPEG (.jpg, .jpeg), WEBP (.webp), GIF (.gif) 2. Fonts: WOFF (.woff), WOFF2 (.woff2), TTF (.ttf), OTF (.otf) ## File Limits 1. Maximum file size: 100 MB per file 2. Maximum total size per Skill: 500 MB (uncompressed) 3. Maximum ZIP upload size: 100 MB (compressed) 4. Number of files: unlimited, provided the total size stays within 500 MB 5. SKILL.md: up to 50,000 characters *** ## **Enable Web Search and Knowledge Search** Skills can access the web and existing knowledge when needed. Both options are set per skill under *Settings*. The search covers only content the respective user has access to in nuwacom. | **Setting** | **Result** | | :- | :- | | Enable Web Search | The skill looks up information on the web when needed. Example: summarizing current market news. | | Enable Knowledge Search | The skill searches the Knowledge Base and uses matching content as context, without files having to be attached individually. Example: drafting offers based on product documentation. | ### **How It Works** 1. Open the *Skills* module. 2. Select the required skill, open the three-dot menu, and select *Edit*. 3. In the edit window, open the three-dot menu and select *Settings*. 4. Switch on *Enable Web Search*, *Enable Knowledge Search*, or both. > [!NOTE] > > When Knowledge Search is active, the skill instruction describes which content serves as context, for example: "Use the current pricing guidelines as context for this offer." *** ## Skill Limits and Activation Up to 40 Skills can be active per user at the same time. Once this limit is reached, newly created Skills are added as inactive by default. Activating a new Skill requires deactivating an existing active Skill first. Shared Skills are inactive by default and can be activated whenever needed. *** ## Call a Skill Typing @ in the chat input applies a Skill directly. This helps when a specific Skill should be applied even though several Skills could be relevant. 1. Type @ followed by the Skill name in the input field, for example @Customer Communication Skill. 2. Add the request in the same message, such as "Write a follow-up email after a customer meeting. Include next steps and keep the tone professional." 3. Send the prompt. nuwacom applies the selected Skill to the request automatically. # Data Analysis Source: https://docs.nuwacom.ai/en/data-analysis Gain insights from your data in seconds. With this feature, you can analyze tables such as Excel, CSV, or Google Sheets directly in nuwacom. The AI uses an integrated Python code interpreter to automatically generate analyses, visualizations, or new files—delivering clear results within seconds, without any manual work. ![Aiappdataanalysis](https://mintcdn.com/nuwacomgmbh/tzlUXQ5D4ImeIx2T/images/aiappdataanalysis.png?fit=max\&auto=format\&n=tzlUXQ5D4ImeIx2T\&q=85\&s=9c079e2d8bd183b865b77ce7a7205332) You can use it, for example, to: * Summarize data and extract key insights * Identify trends, patterns, and outliers * Create charts * Calculate key metrics * Generate new files such as Excel sheets, CSV files, PDFs, or PowerPoint presentations > [!NOTE] > > **What is Python and how does it help with AI-powered data analysis?** > > Python is a programming language that is especially well-suited for processing and analyzing data. It has a simple structure and is therefore widely used in data analysis and AI applications. > > When you analyze data with AI, Python handles the actual work in the background:\ > It reads your files (e.g., Excel or CSV), calculates metrics, identifies patterns, and creates charts or new files if needed. > > Combined with AI, this means for you:\ > Describe as clearly as possible what you want to analyze. Python automatically performs the calculations and analysis, delivering clear and easy-to-understand results within seconds—without requiring any programming. ## **How to analyse data** ![Promptdataanalysis](https://mintcdn.com/nuwacomgmbh/tzlUXQ5D4ImeIx2T/images/promptdataanalysis-1.png?fit=max\&auto=format\&n=tzlUXQ5D4ImeIx2T\&q=85\&s=6018fe5a8cf294ea4f42df53cae52f30) **1. Upload file** Upload your dataset, for example an Excel file with sales data by region, product, and month. **2. Start analysis** Click “Analyze data” or enter a specific request in the chat, for example: *“Analyze revenue by region and product, show me the top performers, and identify any anomalies.”* **💡Tip**: The more specific your request, the better the results. **3. Get results** nuwacom automatically analyzes your data and provides a clear summary, relevant charts (e.g., revenue by region), and actionable insights that you can use directly for decisions or reports. ## **Continue working with your results** You can directly reuse your analysis results and integrate them into your workflows. For example, you can: * Share the entire chat or individual responses with colleagues to ensure transparency or collaborate further * Move relevant content into a project to organize and develop it further * Use the response as a basis for an email, adjust it if needed, and send results quickly * Turn the results into structured documents and continue editing them in the editor, for example for reports or presentations # Documents Source: https://docs.nuwacom.ai/en/documents Create documents directly in nuwacom – fully aligned with your corporate design. Upload your own templates and ensure every document automatically matches your brand. Simply describe the document you want to create and define language, tone, and target audience to get tailored results. This app allows you to quickly create blog articles, concepts, reports, guides, or strategy papers – from the initial idea to the final document. ## **How to create a document** ![Createdoc](https://mintcdn.com/nuwacomgmbh/tzlUXQ5D4ImeIx2T/images/createdoc-1.png?fit=max\&auto=format\&n=tzlUXQ5D4ImeIx2T\&q=85\&s=e56ec10c4b03c4395631daeec865c161) **1. Open chat and start task** Open the chat and start the “Create document” task. **2. Describe the document** Describe the document you want to create in the input field. Example:\ *Create a research report on global e-bike market trends (focus: Europe & Germany), including an executive summary, market overview, key trends, customer demand, as well as opportunities & risks.* **3. Define corporate design & style** Make sure your document fully matches your brand: * Upload your own template or select a template * Define the language * Set the tone of voice * Specify the target audience With a template, nuwacom automatically applies your corporate design, a suitable structure, and a consistent layout. **4. Send prompt** Submit your prompt. **5. Edit document in the editor** nuwacom automatically creates a structured document and opens it directly in the editor on the right. You can immediately edit, refine, and expand your content there. ## **Supported Template Files** You can use the following file formats as templates:\ .doc, .docx, .png, .jpg, .jpeg, .html **Note**\ The maximum file size per upload is 50 MB. ## **Edit document** ![Editdoc](https://mintcdn.com/nuwacomgmbh/tzlUXQ5D4ImeIx2T/images/editdoc.png?fit=max\&auto=format\&n=tzlUXQ5D4ImeIx2T\&q=85\&s=4c9e40b22d5e5a871b00a82b5dbf5026) You can edit and expand your document at any time – in the chat or directly via inline editing in the editor. **Content & structure:** * Add or create new content * Rewrite or optimize text * Adjust structure and formatting directly **Design & layout:** * Customize layout and visual elements * Edit box styles (background, borders, colors) * Adjust border width, style, and radius individually **Images & visuals:** * Upload or generate images * Position and scale images flexibly This allows you to fully align both content and design with your corporate identity. > [!NOTE] > > Save your changes anytime using **“Save version”** in the top right – this keeps all versions traceable and accessible. ## **Manage & organize documents** You can reuse and centrally organize your document at any time. Go to **AI Apps → Documents** to access the document overview. There, you’ll find all created documents and can manage them easily. You can: * **Open and edit documents** – update and refine content anytime * **Share documents** – collect feedback and collaborate with your team * **Move documents into projects** – group content by topic and use it in the right context * **Assign owners** – create clear responsibilities and structure collaboration * **Add tags** – find documents faster and keep everything organized This ensures your content is not isolated, but structured, discoverable, and directly usable within your team. ## **Export document** You can download your document at any time. Available formats: * DOCX * PDF This allows you to use your document outside of nuwacom as well. # Emails Source: https://docs.nuwacom.ai/en/emails Create emails directly with AI in nuwacom. Briefly describe your request or use suggested prompts to get started faster – then run the task “Draft email.” You instantly receive a complete draft with subject and body. Define language, tone, and audience to ensure the email fits your needs. Upload your own email templates if required. Open the draft in the editor and adjust it flexibly – either via inline editing or together with your team. ## **How to create an email** ![Writeemailprompt (1)](https://mintcdn.com/nuwacomgmbh/tzlUXQ5D4ImeIx2T/images/writeemailprompt\(1\).png?fit=max\&auto=format\&n=tzlUXQ5D4ImeIx2T\&q=85\&s=62f9c45e1e351584384784978ecb9cd2) **1. Open chat** Open the chat to start creating your email. **2. Select task** Choose the task “Draft email.” **3. Describe email** Describe the email you want to create.\ Example: *“Create a short follow-up email after a project meeting. Summarize the key points and list the next steps.”* **4. Define setup** Define language, tone, and audience, and attach relevant files if needed. **5. Edit email draft** The email draft is created automatically and opens directly in the editor. Adjust it flexibly or collaborate with your team. **6. Open in Outlook or Gmail & send** Open the email in Outlook or Gmail and send it directly. ### **Tip: Connect Outlook or Gmail** ![Connectoutlookinchat](https://mintcdn.com/nuwacomgmbh/tzlUXQ5D4ImeIx2T/images/connectoutlookinchat.png?fit=max\&auto=format\&n=tzlUXQ5D4ImeIx2T\&q=85\&s=bc78330e0cb8104bf4eec04770c08be8) Connect Outlook or Gmail as a connector in nuwacom. This allows the AI to use emails from your inbox as context and generate replies or new drafts based on existing conversations. How to connect your email account: 1. Open the chat. 2. Click the + icon in the input field. 3. Click *Add connectors.* 4. Select Microsoft Outlook or Gmail under Apps. 5. Click Connect and sign in with your account. Activate the Outlook or Gmail connector, select “Draft email,” and describe your request (e.g. “Reply to the emails I received today.”) – nuwacom searches your inbox and creates relevant reply drafts. ## **Iteratively improve emails** You can refine your email draft at any time via the chat or directly in the editor. Use AI to adjust, shorten, rephrase, or expand content, and edit text manually or together with your team if needed. Example prompts: * *“Make the email more friendly and personal.”* * *“Shorten the email to three paragraphs.”* * *“Add a clear to-do list for the team.”* * *“Make the email more formal for external partners.”* All changes are applied directly in the editor so you can quickly adapt the draft to your style and context. ## **Open and send emails** Once your email is ready, you can open it directly in your email client. Click **Continue** and choose: * **Open in Gmail** * **Open in Outlook** nuwacom automatically transfers the subject, recipients, and content to your email client so you can review and send it there. ## **Continue working with your emails** Via **“Continue”**, you can reuse your email content for other purposes and continue working with it: * **Create document** – Turn the email into a structured document to edit, share, or use in projects * **Create note** – Save key points or ideas as a note and expand them further * **Start new chat** – Use the email as context to analyze, rewrite, or develop content You can also: * download it as a **DOCX file** * download it as a **PDF** * **share it with your team** * **move it to a project** This keeps your content organized and reusable at any time. *** ## FAQ > **How can I use Outlook or Gmail in nuwacom?** > > Connect your email provider like Outlook or Gmail to make your emails available directly in nuwacom. > > The AI can then: > > * search emails and find relevant content > * automatically summarize content > * create suitable reply drafts > > This allows agents to access information from your inbox and actively support you with tasks like replying to emails or reusing content from existing conversations. > **Are my data from connected tools used for AI training?** > > No. nuwacom does not use your company data to train AI models. As a GDPR-compliant data processor, nuwacom processes your data only according to your instructions and contractual agreements. > **How does nuwacom ensure data protection when connecting Outlook or Gmail?** > > When connecting email accounts (Outlook or Gmail) with nuwacom, the following data protection principles apply: > > * **Minimal data access:** nuwacom only accesses emails when you explicitly instruct it (e.g. "Search for emails from X" or "Send an email to…") > * **No permanent storage:** email content is not stored permanently in nuwacom but processed temporarily > * **Encrypted transmission:** all communication is secured via encrypted connections (TLS 1.2/1.3) > * **Role-based access control (RBAC):** you can define exactly which users have access to connected email accounts > * **Audit logs:** all security-relevant activities, including data access, are fully logged and available for audits > > Authentication is handled via secure OAuth mechanisms. The connection can be revoked at any time. # Search Source: https://docs.nuwacom.ai/en/enterprise-search Quickly access information and generate content from your data The module offers you and your team an efficient way to quickly and accurately find the information you need. Simply enter a search term into the search bar—nuwacom scans your chats, documents, and sources in the knowledge database within seconds. The result is a precise and speedy delivery of the necessary information. From the search results, you can directly generate new content, such as creating summaries or generating internal newsletters. > **Intelligent Search Function** > > The module uses advanced algorithms to better understand your queries and maximize the relevance of search results. > **Trusted Information** > > The search results are continuously updated, ensuring you always have the most current data available. This way, you always stay up to date. > **Targeted Insights** > > The existing permissions of your data sources are considered, so you only see the information that is intended for you. ### **Watch the video to see how nuwacom connects to the knowledge of your organization – guided by Nuwa.** [YouTube video player](https://www.youtube.com/embed/GyPSRXMAQ4s?si=JlKKAgBvgz1852Um) *** ## Navigate, Filter, and Sort Search Results Navigate to the "Search" module and enter your search term in the search bar. You will receive matching search results with relevant information and data from your chats, documents, and sources in your knowledge database. * Use the clear source view to filter by your preferred source or select a specific date range to further narrow down the search results. ![Bildschirmfoto 2024-12-12 um 10.35.17.png](https://mintcdn.com/nuwacomgmbh/QmokFyROSwNjqeyr/images/Bildschirmfoto2024-12-12um10.35.17.png?fit=max\&auto=format\&n=QmokFyROSwNjqeyr\&q=85\&s=ea51e60e37c7bf1a6fb2cf2245bfc98d) * Sort the results by relevance or creation date to quickly find the most important information for you. ![Bildschirmfoto 2024-12-12 um 10.37.08.png](https://mintcdn.com/nuwacomgmbh/QmokFyROSwNjqeyr/images/Bildschirmfoto2024-12-12um10.37.08.png?fit=max\&auto=format\&n=QmokFyROSwNjqeyr\&q=85\&s=65e9340aca47021215a2c6fb3be0db4f) ## **From finding information to creating new content** Open a new chat from a result and let the AI create a summary. Alternatively, you can create a content to generate documents like an internal newsletter. ![Bildschirmfoto 2024-12-12 um 09.51.17.png](https://mintcdn.com/nuwacomgmbh/QmokFyROSwNjqeyr/images/Bildschirmfoto2024-12-12um09.51.17.png?fit=max\&auto=format\&n=QmokFyROSwNjqeyr\&q=85\&s=5ca9e83a270ff506be515f3e94979d46) # How to write effective prompts Source: https://docs.nuwacom.ai/en/generateprompts Guide to writing effective AI prompts Prompt engineering refers to the ability to design prompts in such a way that they generate the desired output from AI models. A well-formulated instruction can significantly improve the quality and relevance of the generated content. In this article, we summarize the most effective prompt frameworks to achieve the best possible results in content creation. > **[Generate Prompts](/product/prompts#how-to-create-prompts)** > > nuwacom supports you with an integrated function for **automatic prompt generation**. This allows you to get started quickly and concentrate on your core tasks. You can find more information on this in our article, which guides you step by step through the use of this function. *** ## **Prompt Frameworks** ### **E.R.A (Expectation, Role, Action)** * Expectation: Clearly define what you expect from the AI. For example: "Create a LinkedIn post highlighting the benefits of our new AI platform." * Role: Give the AI a specific role. For example: "Write from the perspective of an HR manager." * Action: Describe the specific action you want the AI to perform. For example: "Explain how these policies improve employee productivity and well-being." **Example prompt:**\ *"Create a LinkedIn post highlighting key policies and best practices for remote work. Write from the perspective of an HR manager and explain how these policies improve employee productivity and well-being."* ### **T.A.G (Task, Action, Goal)** * Task: Describe the task to be completed. For example: "Create a LinkedIn post." * Action: Specify what action you want the AI to perform. For example: "Explain how these policies improve employee productivity and well-being." * Goal: Define the goal of the post. For example: "The goal is to raise awareness of effective remote work strategies." **Example prompt:**\ *"Create a LinkedIn post highlighting key guidelines and best practices for remote work. Explain how these policies improve employee productivity and well-being. The goal is to raise awareness of effective remote work strategies."* ### **C.A.R.E (Context, Action, Result, Example)** * Context: Specify the context. For example: "Remote work has become the norm in many companies." * Action: Describe the action you want the AI to perform. For example: "Create a LinkedIn post." * Result: Define the desired outcome. For example: "The post should clearly communicate the key benefits and best practices of remote work and inspire readers to implement them in their day-to-day work." * Example: Provide a concrete example to give the AI a clear orientation. A good example shows the desired tonality, structure or depth of the output and helps to fulfill the requirements more precisely. **Example prompt:**\ *"Remote work has become the norm in many companies. Create a LinkedIn post on this topic. The post should clearly communicate the key benefits and best practices of remote work and inspire readers to implement them in their day-to-day work. Use the following LinkedIn post as a guide in terms of structure and tone."* ### **R.I.S.E (Role, Input, Steps, Expectation)** * Role: Define the role of the AI. For example: "Write from the perspective of an HR expert." * Input: Enter the necessary information. For example: "Use the information from our latest white paper." * Steps: Describe the steps you want the AI to take. For example: "Create a LinkedIn post that summarizes the most important findings." * Expectation: Define the expectation. For example: "The post should be informative and appealing." **Example prompt:**\ *"Write from the perspective of an HR expert. Use the information from our latest whitepaper on remote work. Create a LinkedIn post that summarizes the most important guidelines and best practices. The post should be informative and engaging."* ### **T.R.A.C.E (Task, Request, Action, Context, Example)** * Task: Describe the task. For example: "Create a LinkedIn post." * Request: Indicate the specific request. For example: "Highlight key policies and best practices for remote work." * Action: Describe the action. For example: "Explain how these policies improve employee productivity and well-being" * Context: Specify the context. For example: "Remote work has become the norm in many companies." * Example: A concrete example or illustration that facilitates understanding. It shows what the task could look like or how similar tasks have been solved in the past in order to minimize uncertainties. **Example prompt:**\ *"Create a LinkedIn post highlighting key policies and best practices for remote work. Explain how these policies improve employee productivity and well-being. Remote work has become the norm in many organizations. Use the following LinkedIn post as a guide in terms of structure and tone."* *** ## More tips **Clarity and precision:** Formulate your prompt as clearly and precisely as possible. Avoid vague or ambiguous terms. A clear and precise prompt helps the AI to understand exactly what result you expect. Unclear: *"Create a report."*\ Clear: *"Create a monthly report on our brand's social media interactions."* **Relevance of the data:** Make sure that the data used in the prompt is relevant and specific to the desired result. Only relevant data leads to useful content. Irrelevant: *"Write an article about food."*\ Relevant: *"Write an article about the latest trends in vegan nutrition."* **User-friendliness:** Design the prompt so that it is understandable and easy to use for all users. Avoid technical jargon and complicated formulations. Complicated: *"Explain the strategic implications of market segmentation."*\ Simple: *"How do we segment our market to better reach our target group?"* **Result orientation:** Formulate the prompt in such a way that it is aimed at a concrete and usable result. Think about what you want to achieve with the answer. Unclear: *"Tell me about our products."*\ Results-oriented: *"Which of our products achieved the highest sales figures in the last quarter?"* **Sustainability:** Think long-term and sustainable when using the knowledge database. Prompts should be designed in such a way that they are also useful in the future. Short-term: *"What are the current sales figures?"*\ Long-term: *"How do I analyze the sales figures to identify long-term trends?"* **Efficiency:** Use AI to make your work measurably easier. A well-designed prompt saves time and delivers precise results. Inefficient: *"Give me information about our company history."*\ Efficient: *"Create a timeline of the most important milestones in the history of our company."* *** After learning the basics of perfect prompt creation and how to use prompt frameworks, we invite you to use your newly acquired skills in nuwacom to create customized content that meets your specific needs. [**👉 How to create prompts**](/product/prompts) # Glossary Source: https://docs.nuwacom.ai/en/glossary Terms and definitions for a quick introduction to nuwacom ### **Agent** Interacts with the user by first asking about the objective of the task and collecting relevant information. Guides the user step-by-step through processes such as the creation of documents or presentations. Enables interactive customization and remains flexible to respond to change requests. At the end, a finished result is provided that can be saved, shared or further edited. Several agents can be configured for repetitive work processes. ### **AI models** Artificial intelligence models that are trained to generate human-like text. They use machine learning and deep learning to understand and produce speech and are used in various applications such as text processing and automated communication. In nuwacom, you can choose from powerful models such as GPT-4.1, Claude 3.5 Sonnet or Gemini 1.5 Pro (Beta) - ideal for text generation, data analysis or multimodal tasks. ### **API key** A unique code that is used to authenticate access to the platform's API. It enables secure communication between various software applications and the nuwacom platform. ### **Categories** Structure prompts according to specific output formats, which makes it easier to find them when generating content with AI. ### **Chat** This module enables searching through your own data and real-time web research to efficiently gather relevant information. With the integrated AI library, teams can create consistent and efficient content. Users can develop and deploy customized AI agents that are deeply integrated into existing tools. The LLM-agnostic interface promotes seamless interaction with multiple language models. ### **Context** Information that is relevant for understanding and processing data or requests. In the nuwacom platform, context helps improve the relevance and accuracy of generated content by providing additional information that the AI takes into account when analyzing and responding to queries. Context information can include files, documents, or text — for example, PDFs with product details, Word documents with press releases, presentations with marketing strategies, or Excel sheets with customer data. This information enables the AI to create content in the right context or answer questions with greater precision. ### **Conversation starters** Introductory questions or topics that are used to start a conversation with the agents. They help to determine the focus of the interaction and obtain relevant information. ### **Custom prompts** Specific instructions created by users to control the AI platform. They can be saved within the prompt library and used for various tasks and applications. They can only be used by the creator or shared with all users in the workspace. ### **Data connectors** Interfaces that make it possible to connect external data sources to the nuwacom platform. They facilitate the import and integration of data from various systems and applications. ### **Data Source** A collection of information or data sets that are used for analysis and processing by the AI. Data sources can include internal documents, websites or the trained knowledge of the AI model that are relevant for answering questions and creating content. ### **Data Source: Knowledge base​** When this option is enabled, the AI accesses internal knowledge stored in your nuwacom knowledge base. This allows it to deliver well-founded answers based on your specific company knowledge – ideal for internal processes, products, or policies. ### **Data Source: Web** This knowledge source allows the AI to access up-to-date information from the internet. It is particularly useful for finding the latest news, external documentation, or dynamic data. The feature can be enabled or disabled as needed – depending on your specific information requirements. ### **Documents** The Documents module enables you to create content in seconds, fully aligned with your individual standards. With AI-powered features, you can instantly generate ready-to-use formats such as LinkedIn posts and blog articles. The "Ask Anything" function lets you refine content directly within the editor — for example, by turning long texts into bullet points or adjusting the tone and style. Documents can be transferred to a project or downloaded with a single click for further use or sharing. ### **Editor** Tool for creating, editing and formatting content directly within the platform. Supports real-time collaboration so that multiple team members can work on a document at the same time. Versioning allows changes to be tracked and previous versions to be restored. The editor combines AI-supported functions with traditional text editing tools. ### **Filters** They are used to refine and optimise search results. Filters help to quickly identify and utilise the most relevant information by sorting or narrowing down the results according to certain criteria. ### **Forms** Prompts displayed as forms enable structured and repeatable workflows. Defined variables are used to collect all necessary information before the prompt is executed — ensuring that all relevant inputs are provided and the results meet your requirements. ### **Generate** The process of creating content or data by AI. In nuwacom, generation refers to the automatic production of texts, reports or other formats based on prompts and data sources. ### **Instructions** Specifications or guidelines that control how prompts work, defining how user-configured prompts should respond to specific requests and what processes should be carried out. ### Connectors Feature that enable a seamless connection of the nuwacom platform with internal data sources such as Confluence, SharePoint, websites and intranet solutions. Ensure automatic data synchronisation and always up-to-date information within the knowledge database. ### **Intelligent Search** A nuwacom module that allows users to quickly and accurately search for information in chats, documents and the knowledge base. The intelligent search function uses advanced algorithms to optimize the relevance of hits and provides a continuous update of search results. As a result, users can directly generate content, such as summaries or newsletters. ### **Knowledge Base** Central component of the AI platform that contains all relevant documents. These include texts, reports, instructions and other written materials that are used to create content and answer questions. ### **Output** In the platform, the output refers to the generated content that is provided after the processing of data and prompts. This can be an answer to a question or an output format within a project. ### **Private files** A view in the knowledge base of documents that the user has uploaded manually within a chat. These documents are not visible to other users in the workspace. ### **Projects** This module enables the efficient creation and management of multiple pieces of content in different formats at the same time. With the help of AI, brand-compliant and consistent content is generated based on centrally stored contextual information. You can add files that the AI uses to create precise and relevant content. By generating different content formats, you can generate multiple pieces of content in parallel and flexibly edit them in the AI editor. All edited content is automatically saved to prevent data loss and ensure seamless collaboration within the team. ### **Prompts** Instructions or questions that are entered into the input field of a chat to perform specific tasks. Can be customized and saved within a workspace. Prompts can be used in chats, when interacting with agents, in projects, in documents, and in the text editor to obtain precise answers and create content efficiently. ### **Response style** The way in which the AI agent responds to requests. The response style can be customized to control the tone, formality and detail of the generated content. The temperature and top-p parameters can be customized. ### **Responsible users** Users who are responsible for specific content. Responsibilities for documents and projects can be clearly assigned to optimize collaboration and content management. ### **Role** A user's specific position within the platform. Roles define the access rights and responsibilities a user has. They ensure that each user only uses the functions that are relevant to their tasks. ### **Search functionality** A function that enables users to search efficiently for relevant documents and information. Users can enter keywords to search for specific documents, articles or information within the platform. ### **Sharing** The sharing feature lets you easily share prompts, agents, chats, documents or proejects with your team – either by clicking the share button or copying the link. You can share items with specific people or across the entire workspace. This promotes transparency, speeds up handovers, and helps avoid duplicate work. ### **Source** Information on the origin of the data used to generate responses. ### **Tags** Keywords or labels that are assigned to documents or projects to improve their categorization and findability. Facilitate the search and organization of information. ### **Tasks** Enables users to organize workflows efficiently. Results can be achieved quickly by selecting a predefined task from the library. Variables serve as input fields in the task form and record relevant information to ensure precise results. ### **Upload** The option of integrating new documents or data into the platform's knowledge database. This continuously expands the information base and ensures that up-to-date and relevant data is available for analyses and decision-making processes. ### **Variables** Customizable input parameters within a task. They appear as input fields to be filled out in the task form and allow users to enter relevant details. This ensures that all necessary information is recorded and precise results are produced. ### **Version** A specific output of a document. nuwacom enables the traceability of changes and the restoration of previous versions of documents. This ensures the consistency and integrity of content by allowing users to revert to previous versions and document changes transparently. ### **Voices** A nuwacom feature that allows users to analyze and save individual language and writing styles from existing texts and apply them when generating output. This ensures consistent, targeted communication, saves time and improves the quality of texts. ### **Web Crawler** A web crawler is an automated program that searches websites to collect, extract and index information. In nuwacom, the web crawler is used, for example, to synchronize company websites with the knowledge base by capturing relevant content and integrating it into the database. This synchronization ensures that the knowledge database is regularly updated and enables seamless integration of website content. The crawler regularly checks for changes and ensures that the latest information is always available. ### **Workspace** Workspaces or project rooms within the platform that enable teams to organize and manage their work. Users can switch flexibly between different workspace to access different projects with separate knowledge databases. ### **Workspace files** A view in the knowledge database of all available documents stored within a workspace. These documents are accessible to all users who are part of the respective workspace. # Images Source: https://docs.nuwacom.ai/en/images Create professional images directly in nuwacom with just a few words. Use suggested prompts or describe your visual and define image style and color mood. The AI generates matching visuals for presentations, newsletters, or social media. Use these example prompts to get started right away: * *Hero image for a landing page with a diverse team* * *LinkedIn post image announcing a product launch with modern branding* * *Recruiting campaign image announcing open positions* ## **How to create images** ![Createimage](https://mintcdn.com/nuwacomgmbh/tzlUXQ5D4ImeIx2T/images/createimage.png?fit=max\&auto=format\&n=tzlUXQ5D4ImeIx2T\&q=85\&s=4b2e4cd99c3265de0be983a88b917869) **1. Open chat and start task** Open the chat and start the “Generate images” task. **2. Describe the image** Describe your desired image in the input field.\ Example: *Hero image for a landing page with a diverse team* **3. Define style and color mood** Select image style and color mood to guide the result, e.g. photorealistic, illustration, sketch, or comic as well as warm, cool, pastel, or monochrome. **4. Generate image** Start the generation – nuwacom automatically creates a matching image based on your description. ## **How to use your own images as reference** Upload your own images to edit, extend, or adjust them stylistically. The AI uses your image as a reference and applies changes such as style, background, or color mood directly.\ \ Try it with these example prompts: * *“Change the background of this image to a bright office scene in flat design.”* * *“Transform this team photo into an illustration in a sketchnote style.”* * *“Remove the background and replace it with an abstract color surface in yellow and gray.”* ![Imageasreference](https://mintcdn.com/nuwacomgmbh/tzlUXQ5D4ImeIx2T/images/imageasreference.png?fit=max\&auto=format\&n=tzlUXQ5D4ImeIx2T\&q=85\&s=609a780d44824f36fe7a28af77cc3114) **How it works:** **1. Open chat and start task** Open the chat and start the “Generate images” task. **2. Upload reference image** Upload your own image to use as a base or style reference **3. Describe the changes** Describe in the input field how the reference image should be modified.\ Example: “Transform this team photo into an illustration in a sketchnote style.” **4. Define style and color mood** Optionally select image style and color mood to further guide the result, e.g. illustration, sketch, or comic as well as warm, cool, pastel, or monochrome. **5. Generate image** Start the generation – nuwacom automatically edits your reference image based on your description. ## **Continue working with your image** After creation, you can directly reuse your image and integrate it into your existing workflows: * **Download image**\ Download the final image and use it for presentations, social media, or external communication. * **Transfer content to a document**\ Add the image or its description to a document to further develop content, for example for concepts, campaigns, or internal materials. * **Move to a project**\ Assign your image to a project to keep it organized and reuse it together with your team. This ensures your image does not remain isolated but becomes part of your workflows. ## **Share and collaborate** You can share the entire chat, including the generated image, with others. This allows your colleagues to: * view the image directly * understand the original prompt * create adjustments or new variations This is especially helpful for gathering feedback, collaborating on content, and sharing best practices within your team. ## **Models** The following models are available for image generation: * **Imagen 4.0 (EU)** – EU-hosted, suitable for privacy-compliant use * **Imagen 4.0 Fast (EU)** – EU-hosted, faster generation with slightly reduced quality * **Gemini 2.5 Flash Image (Nano Banana) (EU)** – EU-hosted, fast and efficient for simple visuals * **Gemini 3.1 Flash Image (Nano Banana 2) (EU)** – EU-hosted, fast generation with improved quality * **Gemini 3 Pro Image (Nano Banana Pro) (US)** – Global-hosted, high quality for more advanced images * **Gemini 3.1 Flash Image (Nano Banana 2) (US)** – Global-hosted, fast generation with improved quality > [!NOTE] > > By default, the Imagen 4.0 model is used (EU-hosted). With the role 'Admin' you can select a different image model in Workspace Settings> Models. # Introduction to Agents Source: https://docs.nuwacom.ai/en/introduction-to-agents ## Overview nuwacom agents handle recurring and complex tasks, from the initial briefing to a reviewed result. An agent is described in natural language without any code, draws on the shared company context, and covers a broad range of tasks. Admins retain central control over permissions, approvals, and logging. Agents are invoked in the chat through an "@" mention. ## Key Benefits * **Built without code**: An agent is described in natural language, then refined in the builder by selecting skills and connectors. * **Grounded in company context**: Agents draw on shared documents, data sources, skills, and connected apps, so results fit the organization. * **Centrally governed and auditable**: Admins define what an agent may do, when approvals are required, and which actions are logged. ## When to Use An agent pays off as soon as the same role, instruction, action, or knowledge base is needed repeatedly. Instructions, knowledge, and actions are configured once and then remain permanently available, so recurring and complex tasks are completed consistently and reliably. ## Use Cases Agents make teams more productive by covering a broad range of tasks: * **Research**: Information from the web or files is gathered in a structured way. A research agent captures information on a topic systematically and presents it clearly. * **Data analysis**: Data is evaluated and results are prepared. An analysis agent structures large volumes of data, identifies relevant patterns, and summarizes the findings clearly. * **Content creation**: Documents, presentations, and emails are generated. A blog-article agent creates an SEO-optimized draft with a clear structure, keyword integration, and meta description from a keyword and target audience. * **Process execution**: Entire process steps are handled automatically. After a blog article is approved, the agent requests the desired Slack channel and notifies the team about the finished content. ## Internal Agents Most agents are built for use inside the workspace, where instructions, knowledge, actions, and settings are configured before the agent is shared with the team. The entire build runs in nuwacom, from the first draft to sharing across the workspace. A dedicated page covers the full setup and sharing lifecycle. ## External Agents An agent can also be embedded outside the workspace, for example on a website through a widget or an iFrame. This makes it available to people who are not part of the workspace, such as website visitors or customers. A dedicated page covers external publishing, allowed domains, and embedding options. ## FAQ > **When is an agent worthwhile instead of a normal chat?** > > An agent pays off as soon as the same role, instruction, action, or knowledge base is needed repeatedly. Instructions, knowledge, and actions are configured once and then handle recurring and complex tasks from the briefing to a reviewed result. > **How is an agent invoked?** > > An agent is invoked in the chat through an "@" mention. > **Who stays in control of what an agent can do?** > > Admins retain central control over permissions, approvals, and logging, defining what an agent may do, when approvals are required, and which actions are logged. > **What makes an agent's answers company-specific?** > > Agents draw on shared documents, data sources, skills, and connected apps, so results reflect the organization's own context rather than generic information. # Introduction to Design systems Source: https://docs.nuwacom.ai/en/introduction-to-design-systems ## Overview Design systems bring a brand’s colors, typography, logos, graphics and design guidance together in one reusable resource. nuwacom applies this visual foundation when creating slides, documents, apps and other branded output. A Design system can be selected directly in chat or chosen automatically by the AI. When several Design systems are available and none has been selected, the AI asks which one to use. ## Key benefits * **Shared visual direction:** Colors, typography, logos, and design guidance give new content a shared visual direction across supported outputs. * **Less repetition:** An available Design system can be applied in chat without describing the brand in every request. * **Room to refine:** Additional brand materials and edits can extend the Design system as requirements change. * **Distinct brand identities:** Separate Design systems keep different brands or design directions from being combined. ## When to use Design Systems Design systems are useful when nuwacom regularly creates content for the same brand. They also help workspaces that produce content for multiple brands or use more than one visual direction. ## Use cases Design systems can guide presentations, documents and Apps. For example, a presentation can use the brand’s colors and typography, while a document can also draw on its logos and design guidance. ## How it works A design system can be built from three inputs: existing brand materials, such as a style guide or presentation, a website that the AI crawls, or a brand description in chat. The AI uses this input to create a design system that can be reviewed and edited. A design system can also be built manually by adding assets directly in the panel on the right. * **Colors and typography:** The brand palette, fonts and text styles are defined in chat or added directly. Font files can be uploaded separately. * **Logos and graphics:** Individual brand assets can be uploaded or extracted from the provided materials, where available in supported formats. * **Design guidelines:** The usage of these elements is described and refined in chat or edited directly. * **Multiple design systems:** Different brands or design directions can be managed, and each system can be enabled or disabled individually. *** ## FAQ **What content can a Design System be applied to?** Design systems can guide slides, documents, PDFs, PowerPoint and Word files, images and Apps. They guide the appearance of generated content but do not guarantee an exact copy of an existing layout. **Can a Design System be created without brand files or a website?** Yes. A description in chat provides the starting point. The builder asks for basic brand details and builds the Design system from the answers. Its elements can then be refined on the page or in chat. **Why is a brand font not appearing in generated content?** A font name does not always provide the font file needed to render it. nuwacom can retrieve files from Google Fonts and Fontshare automatically; other font files must be uploaded. A standard font is used until the required file is available. **Can two Design Systems be applied at the same time?** No. nuwacom does not mix two Design systems. When several are available and none has been selected, nuwacom asks which one to use. **What is recommended if a Brand Skill is already in use?**\ Create and test a Design system. If it meets expectations, deactivate the Brand Skill. Keep using the Skill if it better supports specific expectations or formats. # Introduction to Skills Source: https://docs.nuwacom.ai/en/introduction-to-skills ## Overview A Skill is a reusable set of instructions that defines how nuwacom responds in a specific context. It can specify the tone, structure, format, terminology, and rules for generating output. Once created, a Skill applies automatically whenever the defined context is relevant, so consistent standards can be reused across different tasks without repeating the same instructions. ## Key Benefits * Writing, structure, and formatting are controlled at the Skill level and applied automatically. * Consistent standards are enforced across reports, emails, summaries, and more. * Multiple Skills can be combined to fully define how nuwacom responds. * Proven formats and workflows become shareable across the team. ## When to Use A Skill applies whenever reliable, repeatable results are needed. When reports, emails, or summaries must follow the same structure and tone every time, a Skill enforces this without rewriting instructions or correcting outputs. The standard is defined once and reused across all work. ## Use Cases Skills apply predefined standards to recurring types of work and make consistent output available across the workspace: * A Customer Communication Skill standardizes the wording, terminology, and structure of external messages. * A Report Skill generates reports with defined sections such as an executive summary, key insights, and next steps. * A Customer Support Skill structures replies with a clear answer, resolution steps, and a consistent support tone. Multiple Skills can be combined when different standards apply to the same output. For example, one Skill can define the structure of a report while another controls its tone and formatting. ## Skills vs. Agents A Skill and an Agent serve different roles and work best in combination. | | Skill | Agent | | :- | :- | :- | | Defines | How nuwacom responds | What nuwacom does | | Focus | Format, tone, structure | End-to-end task execution | | Mode | Passive - Activates automatically or on demand | Active - Can run autonomously | | Think of it as | The rulebook | The executor | Applying a Skill to an Agent combines autonomous task execution with guaranteed output quality. When a behavior should always stay consistent regardless of who initiates it, it belongs in a Skill. ## How It Works When a conversation starts, all active Skills run in the background. nuwacom reads each Skill's description and determines which ones are relevant to the request. If a match is found, the Skill is applied automatically, without selection or mention. For example, the prompt *"Write a follow-up email after a customer meeting. Include next steps and keep the tone professional."* triggers a Customer Communication Skill whose description reads *"Use for customer emails, follow-ups, and external communication,"* and the Skill is applied automatically. *** ## Skills Governance Skills in nuwacom are organized into three categories: *System Skills*, *Personal Skills*, and *Workspace Skills*. Each category defines who owns and manages the Skill. ### System Skills System Skills are platform-level Skills developed and maintained by nuwacom. They are not visible in the Skills management interface and cannot be modified or deactivated by workspace admins or users. They run in the background and support how the Orchestrator operates across different tasks and workflows. ### Personal Skills Personal Skills are custom Skills created by Editors or Admins for their own use. They can be activated or deactivated at any time. ### Workspace Skills Workspace Skills are custom Skills created and managed by Editors or Admins for use within the organization. They can support company-specific processes, terminology, formats, and workflows. *** ## Skills and Policies Policies define what is allowed, including applicable standards and restrictions. Skills define how nuwacom operates within this framework, such as tone, structure, format, and working methods. Policies always take precedence. Skills cannot override or bypass them. *** ## FAQ > **How does nuwacom decide which Skill to apply?** > > nuwacom reads the description of every active Skill and matches it against the request. The description is the activation trigger, so it should clearly state when the Skill applies. > **What happens if several Skills are relevant to one request?** > > All matching Skills can apply together. To enforce one specific Skill, mention it directly with *@Skill name* in the input field. > **Do team members need to install a Skill attached to an Agent?** > > No. A Skill attached to an Agent is active automatically for everyone using that Agent. > **How is a Skill activated for general use?** > > Toggle the Skill on in the Skills overview to make it available in chat at any time. > **Do I need technical knowledge to build a Skill?** > > No. Custom Skills are built through a conversation in the input field, without manual setup. > **What is the difference between Skills and Agents?** > > Agents are the executors. Each Agent is fully configurable: what it does, which knowledge it accesses (for example a product description), which actions are available (for example sending a message), and how it operates, including AI model, response style, and tone of voice. A configured Agent knows what to do, how to execute it, and how to guide through the process. > > Skills are the standards. They define what the output looks like in terms of structure, tone, format, and rules. Skills run in the background and keep outputs consistent and immediately usable. > > The core difference: > > * *Agent:* executes tasks using knowledge, tools, and defined settings. > * *Skill:* shapes the quality and consistency of the output. > > Combining both increases their impact. A configured Agent determines what happens and how it runs, while a Skill ensures the result always meets the defined standards. > **How do Skills and Policies relate to each other, and which takes precedence?** > > Policies define the boundaries. Skills define how work is carried out within those boundaries. When a conflict occurs, the policy always applies. > **When should I use a Skill?** > > A Skill applies whenever reliable, repeatable results are needed. When reports, emails, or summaries follow the same structure and tone every time, a Skill enforces this automatically. The standard is defined once and reused across all work, without rewriting instructions or correcting outputs. > **Can I combine multiple Skills?** > > Yes. Skills can be combined to control different aspects of an output at the same time. One Skill defines the structure of a report while another maintains a consistent tone. This allows full standardization of how results are generated and presented. > **Who can see my Skills?** > > Skills are private by default. Sharing a Skill with the team makes it reusable across the workspace, so everyone works with the same standards, formats, and structures without defining them individually. > **Do I need technical knowledge to create a Skill?** > > No. Skills are created through a simple conversation using *Build with AI*. Describe what is needed in plain language, and nuwacom guides through the process and generates the Skill automatically. > **Who can create and manage Skills?** > > Permissions depend on the user's role. > > * *Reader:* existing Skills can be activated or deactivated for personal use. Creating or modifying Skills is not available. > * *Editor and Admin:* own Skills can be created, edited, and deleted, and Skills shared with edit rights can also be managed. Editing or deleting all Skills across the workspace globally is not available. > **Can I use Skills without creating them?** > > Yes. Once a Skill has been shared, it can be activated and used directly in chat. Existing standards become available without creating new Skills. > **What happens when I edit a shared Skill?** > > Changes apply to everyone with access to that Skill. When a Skill is used across a team, updates immediately affect how results are generated. Aligning changes with the team beforehand is recommended to keep results consistent. # Manage Design systems Source: https://docs.nuwacom.ai/en/manage-design-systems The Design systems page provides a central overview of all Design systems available to the current user. From here, Design Systems can be created, activated, edited, shared and managed. ## Activate or Deactivate Design Systems Each Design system card includes a toggle that controls whether it is used automatically in chats: * **On:** The Design system is applied automatically in chats. * **Off:** The Design system is saved but not applied automatically. ## Design System Actions The three-dot menu on a Design system card provides access to all available actions for that Design system: 1. **Start chat:** Opens a new chat with this Design system already applied. 2. **Edit:** Opens the Design system to adjust its colors, typography, logos, graphics and design language. 3. **Share:** Opens the sharing options to make the Design system available to others. 4. **Delete:** Removes the Design system permanently. This action cannot be undone. The toggle next to the three-dot menu is separate from these actions. It controls whether the Design system is used automatically in chats. ## Share a Design System Design systems can be shared with the entire workspace, specific people or user groups. Sharing permissions determine whether recipients can view or edit them. 1. Open *Design systems*. 2. Select the relevant Design system. 3. Open the three-dot menu on its card. 4. Select *Share*. 5. Select the workspace, people or user groups. 6. Set view or edit permissions. ## Manage Design Systems for a workspace Admins can enable or disable the module for the entire workspace. This workspace setting controls access to the module. 1. Open *Workspace settings*. 2. Select *Modules*. 3. Enable or disable *Design systems* with its toggle. *** ## FAQ **What happens when several Design Systems are enabled?**\ A Design system can be selected on the chat start page or in the resource picker. If only one Design system exists in the workspace, it is applied automatically. If several exist and none is selected, the AI asks which one to use. The AI never mixes two Design systems in one chat. **Who can access a shared Design System?**\ A Design system can be shared with the entire workspace, specific people or user groups. Sharing permissions determine whether recipients can view or edit it. **Who can enable or disable the Design Systems module?**\ Admins can enable or disable it for the entire workspace in Workspacesettings>Modules. # Manage Skills Source: https://docs.nuwacom.ai/en/manage-skills The Skills page provides a central overview of all Skills available in the workspace. From here, Skills can be organized, activated, edited, shared, and managed through a consistent set of controls. *** ### Overview The Skills page lists all Skills accessible to the current user. Four tabs structure the view: * **All:** All Skills in the workspace * **Favorites:** Skills marked as favorites for quick access * **My Skills:** Skills created by the current user * **Shared with me:** Skills that other users have shared with the current account *** ### Activate or Deactivate Skills Each Skill card includes a toggle that controls its availability in chat: * **On:** The Skill is active and available in chat. * **Off:** The Skill is saved but not available in chat. The star icon next to a Skill marks it as a favorite for faster retrieval. *** ### Edit Skills Skills can be updated at any time. Two paths are available: **Via the edit view:** 1. Open a Skill from the list. 2. Edit the description or instructions. 3. Save the changes. **Via chat:** Describe the desired changes directly in a chat session that uses the Skill. **Note:** Changes apply to all future uses of that Skill. For shared Skills, updates take effect for the entire team immediately. *** ### Skill Actions The three-dot button on a Skill card opens the *More actions* menu, which provides access to all available actions for that Skill: * **Pin to sidebar:** Adds the Skill to the sidebar for fast, recurring access. * **Start chat:** Opens a new chat session with this Skill already applied. * **Edit:** Opens the Skill to adjust its description or instructions. * **Change avatar:** Updates the avatar displayed for the Skill in lists and chat. * **Duplicate:** Creates a copy of the Skill as a starting point for a new variant. * **Share:** Opens the sharing options to make the Skill available to others. * **Delete:** Removes the Skill permanently. This action cannot be undone. *** ### Filtering, Searching, and Sorting Skills Three controls are available to narrow down the Skills library view. Filters can be combined. A search term and sort order each apply within the selected tab. **Tab Filter** The tab filter determines which Skills are displayed. Available tabs: All, Favorites, My Skills, Shared with me. **Search** The search field ("Search skills…") filters the grid by Skill name as characters are entered. The search runs within the active tab: a search in the *Shared with me* tab returns only shared Skills. **Sort** The *Last used* menu on the right opens the sort options. The grid can be sorted by: * Last used * Most used * Last created * Last updated * Alphabetically *** ### Share Skills Custom Skills are private to their creator by default. They can be shared with individual users, user groups, or the entire workspace. **To share a Skill:** 1. Open the Skill's edit view and select *Share*, or open the *More actions* menu and select *Share*. 2. Choose who should receive access. 3. Confirm the selection. **Sharing options:** * **Specific users:** Share the Skill with one or more individual users. * **User groups:** Share the Skill with all members of a selected group. * **Entire workspace:** Make the Skill available to everyone in the workspace. When a Skill is shared with the entire workspace, it is active by default for all recipients. Individual users can deactivate a shared Skill if they do not need it. ## Pinning Skills Pinned skills appear in the sidebar for quick access. This is useful for skills used frequently, so they are always reachable without searching. ### How to Pin a Skill 1. Find the skill in the skill library or search results. 2. Open the three-dot menu on the skill card. 3. Select *Pin to sidebar*. To unpin a skill, click *Unpin* in the sidebar. ### Accessing Pinned Skills Pinned skills appear directly in the navigation sidebar. Pinning is personal to each user account and does not affect how the skill appears to other workspace members. *** ## Skill Duplication Duplicating a skill creates an independent copy that can be modified without affecting the original. This is useful for creating variations of an existing skill or using another skill as a starting point. ### How to Duplicate a Skill 1. Find the skill in the skill library. 2. Open the three-dot menu on the skill card. 3. Select *Duplicate*. The copy opens in edit mode. The duplicating user becomes the owner of the new skill. ### Roles and Permissions Admins configure which Skills capabilities are available to Editors and Readers under *Workspace settings > Permissions.* **Skills permissions:** * **Read:** View and use Skills in chat. * **Write:** Create and edit Skills. * **Share:** Share Skills with users, user groups, or the entire workspace. By default, Editors have *Read*, *Write*, and *Share* permissions. Readers have *Read* access only. Admins can adjust each permission independently. **To manage permissions:** 1. Open *Workspace settings > Permissions.* 2. Locate the *Skills* section. 3. Enable or disable the required permissions for each role. **Permissions overview:** | Capability | Reader | Editor | Admin | | :- | :- | :- | :- | | Use shared workspace Skills | Yes | Yes | Yes | | Create personal Skills | No | Yes | Yes | | Share Skills with all workspace users (active by default) | No | Yes | Yes | | Recipients can toggle shared Skills off | — | Yes | Yes | *** # MCP Directory Source: https://docs.nuwacom.ai/en/mcp-directory ## Overview The MCP server directory is a curated list of official, remotely hosted Model Context Protocol servers that connect to nuwacom. Each server is operated by the company that builds the underlying product, which keeps connections reliable, secure, and continuously supported. Once connected, an MCP server makes its tools and resources available across Agents, chats, and workflows in nuwacom. ## Trust and verification Every entry in this directory meets three conditions before it is listed: * **Official source**: operated directly by the company that owns the product. * **Remotely hosted**: no local installation required, connection runs through a URL. * **HTTP transport**: secure HTTP or SSE transport protocols. > **GitHub** > > **Auth:** Auth required > > **Transport:** HTTP > > **Server URL:** `https://api.githubcopilot.com/mcp/` > > **Documentation:** [docs.github.com](https://docs.github.com/en/copilot/how-tos/provide-context/use-mcp/use-the-github-mcp-server) > **Atlassian** > > **Auth:** OAuth > > **Transport:** HTTP > > **Server URL:** `https://mcp.atlassian.com/v1/sse` > > **Documentation:** [support.atlassian.com](https://support.atlassian.com/atlassian-rovo-mcp-server/docs/getting-started-with-the-atlassian-remote-mcp-server/) > **Notion** > > **Auth:** OAuth > > **Transport:** HTTP > > **Server URL:** `https://mcp.notion.com/mcp` > > **Documentation:** [developers.notion.com](https://developers.notion.com/docs/mcp) > **Linear** > > **Auth:** OAuth > > **Transport:** HTTP > > **Server URL:** `https://mcp.linear.app/mcp` > > **Documentation:** [linear.app](https://linear.app/docs/mcp) > **Canva** > > **Auth:** OAuth > > **Transport:** HTTP > > **Server URL:** `https://mcp.canva.com/mcp` > > **Documentation:** [canva.dev](https://www.canva.dev/docs/connect/mcp) > **Asana** > > **Auth:** OAuth > > **Transport:** HTTP > > **Server URL:** `https://mcp.asana.com/mcp` > > **Documentation:** [developers.asana.com](https://developers.asana.com/docs/using-asanas-mcp-server) > **Sentry** > > **Auth:** OAuth > > **Transport:** HTTP > > **Server URL:** `https://mcp.sentry.dev/mcp` > > **Documentation:** [docs.sentry.io](https://docs.sentry.io/product/sentry-mcp/) > **Stripe** > > **Auth:** OAuth > > **Transport:** HTTP > > **Server URL:** `https://mcp.stripe.com/` > > **Documentation:** [docs.stripe.com](https://docs.stripe.com/stripe-apps/mcp) > **Cloudflare** > > **Auth:** OAuth > > **Transport:** HTTP > > **Server URL:** `https://bindings.mcp.cloudflare.com/sse` > > **Documentation:** [developers.cloudflare.com](https://developers.cloudflare.com/agents/mcp) > **PostHog** > > **Auth:** Auth required > > **Transport:** HTTP > > **Server URL:** `https://mcp.posthog.com/sse` > > **Documentation:** [posthog.com](https://posthog.com/docs/model-context-protocol) > **Neon** > > **Auth:** OAuth > > **Transport:** HTTP > > **Server URL:** `https://mcp.neon.tech/mcp` > > **Documentation:** [neon.tech](https://neon.tech/docs/ai/neon-mcp-server) > **Postman** > > **Auth:** Auth required > > **Transport:** HTTP > > **Server URL:** `https://mcp.postman.com/minimal` > > **Documentation:** [github.com/postmanlabs](https://github.com/postmanlabs/postman-mcp-server/blob/main/README.md#remote-server) > **Render** > > **Auth:** Auth required > > **Transport:** HTTP > > **Server URL:** `https://mcp.render.com/mcp` > > **Documentation:** [render.com](https://render.com/docs/mcp-server) > **Semgrep** > > **Auth:** Auth required > > **Transport:** HTTP > > **Server URL:** `https://mcp.semgrep.ai/mcp` > > **Documentation:** [semgrep.dev](https://semgrep.dev/docs/mcp) > **Context7** > > **Auth:** Auth required > > **Transport:** HTTP > > **Server URL:** `https://mcp.context7.com/mcp` > > **Documentation:** [context7.com](https://context7.com/docs/resources/all-clients) > **Mintlify** > > **Auth:** OAuth > > **Transport:** HTTP > > **Server URL:** `https://mcp.mintlify.com` > > **Documentation:** [mintlify.com](https://www.mintlify.com/docs/ai/model-context-protocol) > **Replicate** > > **Auth:** Auth required > > **Transport:** SSE > > **Server URL:** `https://mcp.replicate.com/sse` > > **Documentation:** [replicate.com](https://replicate.com/docs/reference/mcp) > **Hugging Face** > > **Auth:** Auth required > > **Transport:** HTTP > > **Server URL:** `https://hf.co/mcp` > > **Documentation:** [huggingface.co](https://huggingface.co/docs/hub/en/hf-mcp-server) > **Microsoft Learn Docs** > > **Auth:** None > > **Transport:** HTTP > > **Server URL:** `https://learn.microsoft.com/api/mcp` > > **Documentation:** [learn.microsoft.com](https://learn.microsoft.com/en-us/training/support/mcp) > **DeepWiki** > > **Auth:** None > > **Transport:** HTTP > > **Server URL:** `https://mcp.deepwiki.com/mcp` > > **Documentation:** [docs.devin.ai](https://docs.devin.ai/work-with-devin/deepwiki-mcp) > **Astro** > > **Auth:** None > > **Transport:** HTTP > > **Server URL:** `https://mcp.docs.astro.build/mcp` > > **Documentation:** [docs.astro.build](https://docs.astro.build/en/guides/build-with-ai/#astro-docs-mcp-server) > **Browser Use** > > **Auth:** Auth required > > **Transport:** HTTP > > **Server URL:** `https://api.browser-use.com/mcp` > > **Documentation:** [docs.browser-use.com](https://docs.browser-use.com/customize/integrations/mcp-server) > **Superglue** > > **Auth:** Auth required > > **Transport:** HTTP > > **Server URL:** `https://api.superglue.cloud/mcp` > > **Documentation:** [superglue.ai](https://superglue.ai/docs/mcp/using-the-mcp) > **UiPath** > > **Auth:** OAuth > > **Transport:** HTTP > > **Server URL:** `https://mcp.nuwacom.ai/uipath/mcp` > > **Documentation:** [docs.uipath.com](https://docs.uipath.com/) > **DATEV** > > **Auth:** OAuth > > **Transport:** HTTP > > **Server URL:** `https://mcp.nuwacom.ai/datev/mcp` > > **Documentation:** [developer.datev.de](https://developer.datev.de/) > **Canva** > > **Authentication:** OAuth 2.0 (custom Canva integration required) > > **Transport:** HTTP (Streamable) > > **Server URL:** `https://api.canva.com/connect/v1/mcp` > > **Documentation:** [Canva.dev](https://www.canva.dev/docs/apps/mcp/) # Orchestrator Source: https://docs.nuwacom.ai/en/orchestrator Start complex tasks directly in chat with the Orchestrator, without having to select the right Agents, Skills, or tools yourself. Describe your task and what you want to achieve. The Orchestrator automatically detects which building blocks are needed, calls them in the right order, and coordinates the work until the final result is ready. For research, it independently brings in a Research Agent. For a presentation, it uses the right Slides Skill. For meeting briefings, email drafts, or internal information, it can include connected tools such as calendars, email, or knowledge sources. ![Orchestratoren](https://mintcdn.com/nuwacomgmbh/EZbccf0IzaHaONyB/images/orchestratoren.png?fit=max\&auto=format\&n=EZbccf0IzaHaONyB\&q=85\&s=e067fe9bea3f8b677dec8541000b652c) ## **What is the Orchestrator?** Think of it like the conductor of an orchestra: the instruments, meaning Agents, tools, and models, are already available. The Orchestrator ensures that they are used at the right time, in the right order, and in the right combination. The Orchestrator uses various capabilities in the background to help you reach structured results, even for complex tasks: * **Planning mode:** Break complex tasks down into clear steps before execution. If needed, the Orchestrator asks follow-up questions to correctly understand the goal, scope, and desired output. * **Persistent memory:** Use existing context across sessions. This allows recurring tasks and ongoing projects to build on previous information and results. * **Skills system:** Automatically access suitable capabilities and domain-specific knowledge. The Orchestrator loads relevant expertise as soon as it is needed for your task. * **Loop detection:** Avoid repeated steps that do not create progress. If the Orchestrator detects a loop, it adjusts its approach and follows a different path. * **Checkpoints:** Work with automatically saved intermediate states. Review results, reset the conversation state, or explore alternative solution paths. *** ## **When should I use the Orchestrator?** Use the Orchestrator when your task consists of several steps and different sources, tools, or results need to be combined. It is especially useful for complex tasks where planning, research, coordination, and the creation of a finished result work together. For simple tasks, such as a short rewording or translation, you can continue working directly in chat. Suitable examples include: ### **Create a launch campaign** Use the Orchestrator to create multiple marketing assets for a product launch from a single request. **Example prompt:** > *Create a launch campaign for my product. Research the product and create the following assets from it: a positioning paper, a LinkedIn post, 3 visual ad ideas for Instagram, and an email that I can send to my team.* **Result:**\ The Orchestrator researches relevant information about the product, derives a positioning from it, and creates several connected assets for different channels. You receive not just individual text snippets, but a coordinated campaign package with strategic context, social media content, ad ideas, and internal communication. ### **Research a company for Sales** Use the Orchestrator to analyze a company comprehensively and derive a personalized sales approach from it. **Example prompt:** > *Research a company using its website, recent news, LinkedIn presence, job postings, and public leadership profiles. Identify likely business goals, challenges, relevant stakeholders, buying signals, and create a personalized outreach plan for Sales.* **Result:**\ The Orchestrator collects information from various public sources, structures the key insights, and assesses possible business priorities. The result is a sales briefing with potential pain points, relevant contacts, indicators of current buying intent, and a suggestion for personalized outreach. ### **Automate meeting preparation** Use the Orchestrator to create a recurring process for meeting briefings. **Example prompt:** > *Create a workflow that checks my calendar, detects upcoming meetings, researches the respective company and participants online, and prepares a short briefing from this.* **Result:**\ The Orchestrator plans a process that takes calendar information into account, detects relevant meetings, and researches additional information about companies and participants. The result is a compact briefing that supports your preparation before the meeting with the most important information, possible talking points, and relevant context. *** ## **How it works** **1. Open chat** Open a new chat. **2. Activate Orchestrator** Activate the Orchestrator by typing `@` in the chat and selecting it. Alternatively, you can click the Orchestrator button below the chat. **3. Enter prompt** Describe your task as specifically as possible.\ \ *Example prompt: Create a launch campaign for my product. Research the product and create the following assets from it: a positioning paper, a LinkedIn post, 3 visual ad ideas for Instagram, and an email that I can send to my team.* **4. Send prompt** Send your request. **5. Receive response** The Orchestrator plans the task, calls suitable Agents and Skills as needed, carries out the individual steps, and delivers the result directly in the chat. *** ## **How to enable agents** Choose which agents the Orchestrator is allowed to use for orchestration. 1. Navigate to **Account Settings → Orchestrator** 2. Click 'Enable agents' 3. Select which personal agents the Orchestrator is allowed to use for orchestration. A smaller, carefully selected set helps keep the results focused and relevant.\ > [!NOTE] > > **Note:** By default, the Orchestrator uses **all activated nuwacom agents** and the **personal agents you have enabled in Account Settings**. The Orchestrator operates at the user level, not the workspace level. It does not automatically access all agents in the workspace or agents activated by other users. Each user can therefore tailor their agent selection to their individual responsibilities and professional context. *** ## **Access** As an Admin, you define which roles are allowed to use the Orchestrator in **account settings → Permissions**. In the Orchestrator section, you can enable or disable the Read permission for Editors and Readers. Then click Save changes so the settings are applied. > [!NOTE] > > Access to the Orchestrator can depend on your role in the Workspace. If the Orchestrator is not visible in the chat or in your navigation, access may have been restricted by an administrator. In this case, contact your Workspace administration. # Overview Data Security Source: https://docs.nuwacom.ai/en/overviewdatasecurity Last updated: September 21, 2026 # Handling of personal and private data nuwacom is an AI-powered platform that connects a company’s knowledge and makes it usable for day-to-day work. nuwacom provides employees with fast and easy access to documents, meetings, and connected systems. A key benefit is personalization: results, answers, and suggestions are tailored to the respective user and their permissions. nuwacom operates the platform as Software-as-a-Service in EU data centers and sets up each customer as a separate, logically segregated tenant; access to the customer’s data by other customers or unauthorized third parties is excluded. nuwacom uses only what the customer provides as the data basis: data from the systems or connectors that the customer connects, and content that the customer creates directly in nuwacom. nuwacom does not use its own data sources, and the permissions from the connected systems continue to apply unchanged. To ensure the functionality of the platform, nuwacom processes personal and private data. This document describes what data this is, the purpose for which nuwacom processes it, where it is stored, and when nuwacom deletes it. It is intended as a basis for the customer’s internal reviews, data protection impact assessments, and works agreements. *** ## 1. Allocation of roles and responsibility nuwacom acts toward customers as a processor within the meaning of Art. 28 GDPR. The customer remains the controller within the meaning of the GDPR. nuwacom processes personal data exclusively on the customer’s documented instructions and within the framework of the data processing agreement (DPA), the terms of use, and the platform agreement. The customer alone decides which documents, meetings, and information are added to the platform. nuwacom has no control over the identity of the data subjects or over the types of personal data processed. The transparency and information obligations under Art. 13 and 14 GDPR therefore lie with the customer; nuwacom supports the customer with documentation, evidence, and technical functions. > **Data Controller** > > The customer \ > – determines the purposes, content, user groups, and configuration. > **Data Processor** > > nuwacom GmbH\ > Universitätsstraße 3\ > 56070 Koblenz \ > – operates the platform. > **Data Protection Officer** > > heyData GmbH\ > Schützenstraße \ > 10117 Berlin\ > [datenschutz@heydata.eu](mailto:datenschutz@heydata.eu) > **Principle:** nuwacom does not use customer data to train general AI models. nuwacom does not analyze content for its own purposes and does not share it with third parties that are not contractually bound as sub-processors. ## 2. User profiles and access data nuwacom creates a user profile for each authorized person. It is used for authentication, assignment of permissions, and personalization of work results. For this purpose, nuwacom processes: * **Professional contact and profile data** – first and last name, email address, position, department, location, and other required or voluntary profile information * **Login data** – email address and password or the claims transmitted by the customer through the SSO process If the customer uses Single Sign-On and SCIM, nuwacom manages end-user identities via the customer’s identity provider. The customer therefore controls the creation, modification, and deactivation of accounts themselves. ## 3. Content and connected knowledge sources As content, nuwacom processes all data that the customer and its users add to or generate in the platform: documents, files, prompts, configurations, knowledge bases, search queries, chat histories, outputs, agents, skills, workflows, and apps as well as – if activated by the customer – meeting transcripts and summaries. This content may include personal data of the customer’s employees, authorized third parties, and meeting participants. nuwacom ensures that content is displayed only to those users who are authorized to access it. Access follows a role-based permission concept based on the least-privilege principle; for connected systems, the permissions assigned there remain authoritative. ## 4. Usage data and logging For operation, security, and error analysis, nuwacom processes technical usage data such as IP address, device characteristics, access times, and user ID. Access and audit logs are tamper-resistant and make it possible to trace input, changes, and deletion. > **No performance or behavior monitoring by nuwacom.** nuwacom does not analyze usage data to evaluate the behavior or performance of individual employees. Analyses to improve the services are carried out exclusively in anonymized or aggregated form at workspace level – such as usage statistics, error and performance logs, security monitoring, and model routing. None of these analyses access content as content. > > Access by nuwacom employees to customer data occurs only at the customer’s request in a support case and is logged. ## 5. Meeting notetaker If the customer activates the function, nuwacom can transcribe and summarize meetings. Transcripts and summaries are customer data and are stored in the customer’s workspace, where the customer can view, edit, and delete them. * The customer determines which meetings are recorded and which meetings the agent may attend; individual meetings can be excluded and the agent can be removed. * The agent actively informs participants in the meeting about the active note-taking function; nuwacom also provides consent management functionality, including dual-party consent. * The customer is responsible for determining the legal basis for transcripts and for obtaining the required notices and consents. If the customer deletes a transcript or terminates the account, nuwacom permanently removes the transcript from active systems and backups within 30 days. ## 6. Processing by AI models nuwacom forwards requests to the language models approved by the customer. Processing takes place with the providers listed in the sub-processor list; nuwacom excludes permanent storage of content by the model providers. Customer data is not used to train general AI models. Individual AI services offer different operating modes (global or EU-only). nuwacom enables the customer to restrict processing entirely to the EEA: in the workspace settings, only models hosted in the EEA can be permitted and all other models and regions can be disabled. The customer’s administrators configure this centrally; it applies bindingly to all regular users – models that have not been approved are not available for selection in the interface. Initial setup is carried out during onboarding together with the responsible Customer Success Manager; nuwacom confirms the configured settings in writing upon request. As long as no EU-only mode is configured, prompts and outputs may be processed outside the EU. If such processing occurs, EU Standard Contractual Clauses, adequacy decisions, or the EU-US Data Privacy Framework apply. > **Privacy Mode:** nuwacom anonymizes personal data upon request before it is passed to a model and restores the context for users after processing. The result remains fully readable, while model processing takes place without clear-text data. The function is currently in open beta. ## 7. Storage location, hosting, and sub-processors nuwacom stores customer data at rest exclusively within the EU. Backups are performed zone-redundantly within the same region; backup or recovery data is not outsourced across regions. The storage location is contractually anchored through the hosting option selected in the order form. | Hosting option | Configuration | | :- | :- | | Standard | Microsoft Azure, data centers within the EU | | Germany / STACKIT | At the customer’s express choice; operation in the eu01 region (Germany) | nuwacom engages sub-processors only to the extent necessary and contractually obligates them to at least equivalent confidentiality and data protection duties. nuwacom publishes the current list of sub-processors, including location, service, and transfer basis, in the [Trust Center](https://trust.nuwacom.ai). nuwacom notifies the customer of changes; the customer may object within 15 days on data protection grounds. ## 8. Special categories of personal data nuwacom excludes the processing of special categories of personal data under Art. 9(1) GDPR by default. The data processing agreement expressly states that the services are not intended for the processing of sensitive data. For customers who are subject to professional secrecy obligations under Section 203 of the German Criminal Code (StGB), nuwacom offers a separate supplementary agreement. ## 9. Deletion, return, and retention periods nuwacom deletes personal data as soon as there is no longer a business purpose and returns it at the customer’s request via the provided export interfaces. nuwacom provides a deletion log upon request. Data carriers are destroyed in accordance with DIN 66399 and with reference to NIST SP 800-88 Rev. 1. | Data category | Retention | | :- | :- | | Active customer data | 30 days rolling; monthly snapshots for 12 months | | Backup copies | 30–90 days; complete destruction no later than 90 days after the end of the contract | | Audit logs / access logs | 30–90 days, then automatic deletion | | Meeting transcripts | According to the customer’s configuration; after deletion, removal from active systems and backups within 30 days | | Customer account after contract end | Immediate deactivation; complete deletion within 90 days | ## 10. Technical and organizational measures * **Tenant separation** – dedicated, logically segregated tenant for each customer; row-level security prevents data from being mixed * **Encryption** – TLS 1.2/1.3 in transit, AES-256 at rest * **Access control** – role-based permission concept, MFA for remote access, review at least annually * **Monitoring** – log management, intrusion detection, annual penetration tests, quarterly scans * **Personnel** – confidentiality commitment and mandatory annual awareness training nuwacom is **ISO/IEC 27001:2022** **certified** (SOCOTEC, October 27, 2025). ## 11. Data subject rights and security incidents nuwacom promptly forwards requests from data subjects to the customer as the controller and supports the customer to the technically possible extent in responding to them. nuwacom informs the customer of a personal data breach without undue delay and supports the customer in fulfilling its obligations under Art. 33 and 34 GDPR. ## 12. Further documents > **[Trust Center](https://trust.nuwacom.ai)** > > Certifications, technical and organizational measures, subprocessors > **[Data Processing Agreement](https://nuwacom.com/en/data-processing-agreement)** > > Information on data processing > **[Privacy Policy](https://nuwacom.com/en/privacy-policy)** > > Information on data privacy # Plans Source: https://docs.nuwacom.ai/en/plans ## Overview nuwacom is based on a model of Plans and Seats. Two Plans are available: Business and Enterprise. Within each Plan, every user is assigned one of three Seat-Types: Standard, Premium, or Power. The Seat-Type determines the included AI usage and the price, while the feature scope within a Plan remains the same regardless of the Seat-Type. When demand is higher, Additional AI usage can be booked on top. *** ## Plans nuwacom provides two Plans. Both include the full feature scope and differ in deployment, governance, and integration depth. | Plan | Suitable for | Hosting | Pricing | Booking | | :- | :- | :- | :- | :- | | **Business** | Teams that want to work with AI quickly | Europe on Azure | From €25 per user / month | Directly in the workspace via the Upgrade button | | **Enterprise** | Companies with company-wide AI adoption and maximum control | Europe on Azure, optionally Germany on STACKIT | Individual, from 20 users | Via an AI expert | A detailed feature and pricing overview is available [here](https://nuwacom.com/en/pricing). To unlock advanced capabilities, switch to Enterprise and [talk to an AI expert](https://insights.nuwacom.ai/meetings/robert-ivanov/demo?uuid=5b6d58fb-481d-46b1-8818-416c45ac9fb7). *** ## Seats Three Seat-Types are available within each Plan. The factor describes the included AI usage relative to the Standard Seat. | Seat-Type | Suitable for | Included AI usage | Price per user / month | | :- | :- | :- | :- | | **Standard** | Most employees in regular daily work: chat, knowledge search, summaries, meetings, simple documents | 1x | €25 | | **Premium** | Users with regularly more intensive AI usage: longer files, many meetings, frequent documents, first workflows | 2x | €49 | | **Power** | Very active users, AI Champions, and Builders: many files, tests, agents, complex tasks, apps, and workflows | 5x | € 99 | *** ## Changing a Seat A user's Seat is changed by admins in the settings under "AI usage per user". 1. Open Settings > Plan & Usage > AI usage per user. 2. Go to the row of the relevant user. 3. Open the dropdown in the "Seat" column. 4. Select the new Seat-Type and confirm the change. The new Seat-Type and the included AI usage apply immediately, and the Seat price adjusts accordingly. If no Seat has been assigned to the user yet, a user can be added and a free Seat assigned via "+ Invite user". *** ## When to upgrade is recommended A user's usage behavior indicates the appropriate Seat-Type. An upgrade to Premium or Power is advisable when: * the included AI usage is regularly exhausted. * chat regularly switches to auto mode. * agents, workflows, or apps are used intensively. * Additional AI usage is needed on an ongoing basis. *** ## What happens when AI Usage is exhausted When a Seat's monthly AI usage is exhausted, two automatic responses take effect. No unplanned costs for higher-tier models arise in the process. * Chat automatically switches to Auto mode. * External chatbots and workflows are paused. Chatbots and workflows become available again as soon as Additional AI usage is booked, the Seat is upgraded, or a new usage period begins. *** ## Additional AI usage Additional AI usage is available independently of the Seat and can be flexibly assigned to individual users. Assignment takes place in the "Additional" column under "Usage per user". Additional AI usage: * is available independently of the Seat. * can be flexibly assigned to individual users. * is available in annual packages: €1,500, €3,000, €6,000, €12,000, €24,000, €48,000, or individual. * does not expire at the end of the month. [Learn more](/ai-usage) *** ## FAQ > **Does a different Seat-Type change the feature scope?** > > No. The feature scope remains the same within a Plan regardless of the Seat-Type. The Seat-Type only determines the included AI usage and the price. > **Can I assign more AI usage to a user without changing the Seat-Type?** > > Yes. Additional AI usage can be booked via "+ Buy more". Assignment then takes place in the "Additional" column. Additional AI usage is independent of the Seat and does not expire at the end of the month. > **How do users see their AI usage?** > > Users see how much AI usage is left as a percentage at the bottom left of their account, for example "62% available". Admins see additional details in the settings: > > * consumption per user > * the remaining Additional AI usage in total and per user > * warnings at high usage > **When does a Seat change take effect?** > > Immediately. The new Seat-Type and the included AI usage are active right away, and the price adjusts accordingly. > **Who can manage Plans and Seats?** > > The Plan & Usage area is reserved exclusively for admins. # Policy Hierarchy Source: https://docs.nuwacom.ai/en/policy-hierarchy ## Overview When a user submits a prompt in nuwacom Chat or with an Agent, the system applies multiple layers of guidelines and knowledge sources. This article explains the order of precedence and how conflicts are resolved. *** ## Order of Precedence When a prompt is submitted, the system applies elements in this order (highest to lowest priority): 1. **Prompt/Agent Instruction** (System Instruction): Custom Agent instructions or default system behavior. If chatting with an agent and the instructions are available, it takes it. If not, it uses the default nuwacom prompt. 2. **Tone of Voice:** Writing style guidelines are applied after system instructions. 3. **Workspace Guideline** (Company Guidelines): Workspace-level rules and policies. This applies to all users in the workspace. 4. **Project Guideline:** Project-specific rules and constraints. More specific than workspace guidelines 5. **Private Memory:** User-specific stored information. Stores personal preferences and context. 6. **Context Knowledge** (Attached Files/Documents): Files attached to Chat or Agent. This often has the highest priority for context. 7. **Knowledge Base:** When the knowledge-base tool is enabled on the frontend and there are matching info in the knowledge-base, it is then added. 8. **Web** (External Information): Results from external web search. *** ## Conflict Resolution Some elements can override earlier ones when they conflict. **Example:** * Workspace Guideline: "Always use formal language" * Tone of Voice: "Use casual, friendly tone" * Result: Tone of Voice takes precedence ### Information Sources: Tool-Provided Data Wins When gathering information, those provided by tools (from Context Knowledge, Knowledge Base, or Web) always takes priority over the model's training data. *** ## Exceptions and Special Cases If specific documents are attached to an Agent, the knowledge base search tool may be disabled to avoid redundancy. *** ## Agents vs. Direct Chat Interactions ### Agents * Can have custom System Instructions that override default behavior * Context is merged with agent-specific knowledge first, then request context * Some tools may be conditionally enabled based on agent configuration ### Direct Chat * Uses default system instructions unless an Agent is selected * All workspace and project guidelines apply *** ## Why This Matters This hierarchy ensures: * Consistent behavior: System instructions set the foundation * Flexibility: Workspace and project guidelines add organizational constraints * Personalization: Private Memory provides user-specific context * Accuracy: Information retrieval tools gather current, relevant data in a defined order The system prioritizes user instructions and tool-provided information while respecting organizational guidelines and maintaining a consistent tone. *** ## Best Practices 1. Set workspace guidelines for organization-wide policies 2. Use project guidelines for project-specific requirements 3. Define tone of voice at the workspace or project level for consistency 4. Attach specific documents when you need precise information 5. Enable knowledge base search for broader workspace knowledge 6. Use web search for real-time or external information needs # Privacy Mode Source: https://docs.nuwacom.ai/en/privacymode ## Overview ![Privacymodenenglish](https://mintcdn.com/nuwacomgmbh/gIBh18us90pfbdbm/images/privacymodenenglish-1.png?fit=max\&auto=format\&n=gIBh18us90pfbdbm\&q=85\&s=7708a2f62cd00f3b11f27cf79c3011ec) Privacy Mode detects personally identifiable and sensitive information (PII) in chat inputs and uploaded documents, then replaces it with neutral placeholders before any content is sent to an external AI model. This reduces the risk that personal data is disclosed to external AI providers or stored in log and processing data. The AI still receives enough context to answer the request, while the original values remain protected inside nuwacom. ## Key Benefits * **Anonymize data before processing:** Recognized information — such as names, contact details, customer and employee data, financial information, contract details, and other confidential business data — is replaced with neutral placeholders before it reaches the AI model. * **Protect prompts and files:** Anonymization applies to both chat input and uploaded documents. * **Stay informed:** Anonymized files are clearly labeled, and protected values appear as placeholders in conversations. * **View original information when needed:** Hover over a placeholder or switch between the anonymized and original display within nuwacom. > [!NOTE] > > Enterprise customers can try Privacy Mode at no additional cost until the end of the year. To enable it for your organization, contact your Customer Success Manager. ## When to Use Privacy Mode is designed for organizations with elevated requirements for data protection, information security, and compliance. Typical contexts include healthcare, finance and insurance, public administration, legal and consulting services, HR, and other regulated or data-sensitive areas. Use it whenever a message or document contains personal data that should not be exposed to an external AI provider. ## How It Works With Privacy Mode active, processing runs through several steps: 1. **Input or document upload**: A message is entered in nuwacom, or a document is uploaded for processing. 2. **Automated detection**: The content is scanned for personal and sensitive information, including names, email addresses, phone numbers, identification numbers, and further configurable attributes. 3. **Placeholder replacement**: Detected values are replaced by unique, neutral placeholders, for example Maria Schneider becomes \[Person] and [maria@example.com](mailto:maria@example.com) becomes \[EMAIL]. Hover to view the original value. 4. **Protected transmission**: Only the anonymized version is sent to the external AI model, which processes the request without access to the original values. 5. **Restoration inside nuwacom**: Where required, nuwacom holds the mapping between placeholders and original values in protected form and restores them in the generated response for authorized users. A locally operated model is used for detection, so that detection takes place within the controlled system environment before content is passed to an external AI model. ## How to enable Privacy Mode Admins manage Privacy Mode in the Workspace Settings, on the Privacy Mode page under AI Guardrails. They have two options: enforce it so it stays enabled for all users at all times, or leave it up to each user to enable or disable it as needed. 1. Open Workspace Settings and select *Privacy mode* under AI Guardrails. 2. Use the *Enforce Privacy Mode for all users* toggle to define the workspace behavior. When enabled, Privacy Mode is always on for everyone in this workspace and cannot be turned off in the chat. When disabled, each user decides in the chat whether to enable Privacy Mode for their messages. ## Using Privacy Mode in the Chat The Privacy Mode control is located next to the chat input and is marked with a shield icon. This example shows how to process a customer complaint containing sensitive data. 1. **Enable Privacy Mode:** Activate the shield icon next to the chat input. If enforced by an admin, Privacy Mode is always active. 2. **Attach the document:** Upload the customer complaint using the attachment button. 3. **Send the prompt:** Enter a request to process the attached document and send it. 4. **Review the anonymized content:** Sensitive data is replaced with placeholders such as \[NAME=…], \[DATE=…], \[ID=…], or \[IBAN=…]. Hover over a placeholder to view the original value. 5. **Disable anonymization:** Turn off the anonymization toggle to replace the placeholders with the original names, dates, and numbers. This view is visible only to you. This way, the customer complaint is processed by the AI without exposing personal or sensitive data. Names, dates, IDs, and financial information stay within the organization — the external LLM only sees anonymized content. *** ## FAQ > **Does Privacy Mode make my data fully anonymous?** > > Yes. PII Scrubbing anonymizes personal data before it is sent to the LLM. Recognized values are replaced with neutral placeholders, so external AI models process only the anonymized version and do not receive the original personal information.  > > The mapping between placeholders and original values remains securely within nuwacom, allowing authorized users to view or restore the original data in the nuwacom interface when needed. > **Can the AI still understand my request if data is masked?** > > Yes. Placeholders remain consistent within a request, so the model keeps the context without receiving the real values. > **Does Privacy Mode also protect log data?** > > Yes. The anonymized version is used in system logs and stored processing data to avoid unnecessary exposure. > **Can users turn Privacy Mode off?** > > Only if enforcement is disabled in the workspace. When enforcement is enabled, Privacy Mode stays on for everyone and cannot be switched off in the chat. # Advanced Source: https://docs.nuwacom.ai/en/product/advanced **Explore what’s possible and learn how to make the most of nuwacom’s power features.** Take your workspace to the next level with nuwacom’s advanced capabilities. From building custom agents and managing your knowledge base to setting up connectors — these features give you full control to scale how your teams create, collaborate, and communicate with AI. *** ### **1. Organize Workspaces**  Think of workspaces as different rooms in a house — each private, organized, and decorated for its purpose. Marketing might have a space full of campaign briefs and content drafts, HR another with policies and onboarding guides. Sensitive projects like M\&A or R\&D can stay completely separate, with access tightly controlled.  Inside each workspace, you can also define your **brand guardrails and styles**. This ensures that every piece of content — whether for support, sales, marketing, or legal — follows the same rules and tone. Guardrails keep your communication consistent, professional, and always on-brand.  ![Createworkspace.png](https://mintcdn.com/nuwacomgmbh/OOCwOREF0Ivt7z3S/images/Createworkspace.png?fit=max\&auto=format\&n=OOCwOREF0Ivt7z3S\&q=85\&s=b5709144b36fb230a0f9f5459bf16d32) **What to do inside nuwacom:**  * Go to **Workspace Settings**.  * Click + Create Workspace, enter a short workspace ID and give it a clear name (e.g. *Marketing* or *HR*).  * Open **Workspace Settings > General**.  * Add a recognizable logo (e.g. company, department, or project) to make your workspace easy to identify. * Open **Workspace Settings > AI Guardrails**.  * Define personality traits (e.g. professional, clear, approachable) for this workspace, set tones per context, add forbidden words, and upload style examples.  👉 This way, your knowledge stays structured, your workspaces stay organized, and your brand voice is protected at scale. *** ### **2. Bring Your Team into the Workspace**  ![Inviteusers.png](https://mintcdn.com/nuwacomgmbh/OOCwOREF0Ivt7z3S/images/Inviteusers.png?fit=max\&auto=format\&n=OOCwOREF0Ivt7z3S\&q=85\&s=e53927626ff1df67ecb9ee2b1426948d) In your workspace settings, open **Users** to bring your team together and collaborate effectively:  * **Invite Users**: Add colleagues with a quick email invitation so everyone can work in the same space.  * **Roles and Permissions**: Give each team member the right level of access — *Admin, Editor, or Viewer* — so everyone can contribute without confusion.  👉 Managing users ensures your whole team can collaborate smoothly and stay aligned in one workspace. [*Learn more.*](/product/usermanagement#roles-and-permissions) *** ### **3. Manage Your Knowledge Base**  ![Knowledgebaseen](https://mintcdn.com/nuwacomgmbh/dP_XChvbzqsyD4Jb/images/knowledgebaseen-1.png?fit=max\&auto=format\&n=dP_XChvbzqsyD4Jb\&q=85\&s=1af881be9e3f507be405d243856d98e3) The Knowledge Base is the **central library of truth**. It’s where your most important documents live so the AI can search, reference, and answer questions with confidence. Without it, AI would just be guessing. With it, answers become precise and backed by your company’s data.  **What to do inside nuwacom:**  * Navigate to **Knowledge Base** in your workspace.  * Upload essential documents (manuals, reports, guidelines).  * Organize them into folders for clarity.  * Adjust [permissions](/product/usermanagement#roles-and-permissions) so only the right people can access sensitive material.  👉 Once set up, you can ask: “What’s our vacation policy?” and the AI will answer directly from HR docs. *** ### **4. Connect Data Sources** Your company’s knowledge is scattered across different tools. With connectors, you don’t have to move it — nuwacom simply syncs it. That way, when you ask a question, the AI can pull from your OneDrive project docs or Google Drive presentations, all at once.  ![Chatselectsourceen](https://mintcdn.com/nuwacomgmbh/dP_XChvbzqsyD4Jb/images/chatselectsourceen-2.png?fit=max\&auto=format\&n=dP_XChvbzqsyD4Jb\&q=85\&s=5270d3a10e004a1fa3c4bd8b48a83d6b) **What to do inside nuwacom:**  * Go to **Chat > Add connectors**  * Select an app (OneDrive or Google Drive).  * Click Connect. A window opens where you can allow access and sign in with your account to finish the connection. 👉 *Once connected, click\_+ and Add from apps \_ to choose a file (e.g., Market\_Analysis\_2025.pdf), then type: “Read this market analysis and create three key takeaways we can use in our next campaign brief.” The AI will pull and compile them for you.* > [!NOTE] > > Some knowledge doesn’t sit in tools or files — it lives on websites, intranets, or product portals. Contact us to learn more about our customized connector solutions. *** ### **5. Create and Manage Agents**  Agents are your **AI colleagues** — specialized, always available, and tailored to your business. Instead of explaining the same instructions every time, you configure an agent once, and it becomes an expert in that domain. Picture an “HR Policy Bot” that answers employee questions, or a “Crisis Communicator” that instantly drafts press statements in your brand voice.  **What to do inside nuwacom:**  * Open **Agents**.  * Click **Create Agent**. Describe what you need or what the agent should do, and nuwacom will build the agent for you. E.g. *You are a Research Analyst. Your job is to read through reports, extract the most relevant information, and provide a clear, concise summary.*  * **Add Files**: select the docs, data sources, or style it should use.  * Save your changes, try a few sample queries, and adjust the instruction if needed.  * Share it with your team.  💡 *Example: A “Sales Enablement Agent” can pull case studies, write proposal drafts, and ensure messaging matches your pitch deck.*  🎥 **Watch the full tutorial to learn how to build your own agent in nuwacom:** [YouTube video player](https://www.youtube.com/embed/wEehlfHx_NU?si=UPk-qJ0YHOlId679) *** ### **6. Integrate Tools** nuwacom already connects to 100+ tools, from Outlook to Slack. With nuwacom, you can connect your favorite business tools directly to your agents — making it easy to automate workflows and streamline your daily tasks. ![Integrations.png](https://mintcdn.com/nuwacomgmbh/OOCwOREF0Ivt7z3S/images/Integrations.png?fit=max\&auto=format\&n=OOCwOREF0Ivt7z3S\&q=85\&s=8619d2db3f2d350372b66f43601dac02) **What to do inside nuwacom:**  * Go to **Workspace Settings > Connectors**.  * Explore the tool-catalog.  * Connect standard apps like Sharepoint, Confluence, Slack or MCP-Servers.   * Authenticate with your account credentials.  * Choose which folders or workspaces to sync.  * Bring files and information from connected tools (such as Google Drive, SharePoint, and more) into chats as knowledge, and enable them to empower agents with actions. 💡 *This transforms nuwacom from a passive assistant into an active member of your toolchain.* [*Learn more.*](/product/connectors) > [!NOTE] > > For custom systems, contact us to explore tailored connector solutions for your organization. *** ### **7. Manage LLMs**  Not all AI models are the same. Some are fast and lightweight; others are creative but slower. nuwacom lets you choose which Large Language Models (LLMs) are available in your workspace, so your team always uses the right tool for the job.  **What to do inside nuwacom:**  * Open **Workspace Settings > Models**.  * Select your **default model** (e.g., GPT-5 for depth).  * Enable additional models (Mistral, Cohere, etc.) for specialized tasks.  * Decide which workspaces or teams can access which models.  👉 [*This flexibility ensures you balance performance, cost, and quality. Learn more.*](/resources/aimodels) *** ### **8. Build on Top of Our API**  For teams with developers, nuwacom opens up even more power. The API lets you pull data in or out, extend features, or embed AI into your own tools.  **What to do:**  * Access **API documentation** at [docs.nuwacom.com](/api-reference/ai/list-all-ai-models).  * Generate an API key in **Admin Settings**.  * Use endpoints to:  * Query your Knowledge Base.  * Trigger workflows or prompts programmatically.  * Sync with your internal systems.  💡 *Example: A developer could build a Slackbot powered by nuwacom’s knowledge, answering team questions instantly.*  *** ### 9. Automate Workflows   Today, you can chain prompts, agents, and connectors. Soon, you’ll be able to visually design full **workflows** that run automatically — reducing manual steps.  ![Buildworkflowen](https://mintcdn.com/nuwacomgmbh/dP_XChvbzqsyD4Jb/images/buildworkflowen-2.png?fit=max\&auto=format\&n=dP_XChvbzqsyD4Jb\&q=85\&s=1924ab83710ed6904e38d1b0aa286601) **What it will look like:**  * Define a trigger (e.g., “A new doc is uploaded” or “Slack message received”).  * Add steps (summarize → create Jira task → notify Slack).  * Insert conditions and human approvals for compliance.  💡 *Example: Every time a contract is uploaded, nuwacom could auto-generate a risk summary, create a legal review task, and alert the legal team in Teams.*   👉 [Explore workflows.](/product/workflows) # Apps Source: https://docs.nuwacom.ai/en/product/apps ## **Overview** The Interactive App Builder enables the creation of fully functional apps, dashboards, and internal tools directly through conversational input. The system generates real, working apps within the workspace based on plain-text descriptions, allowing for iterative, chat-based refinement. The system supports the integration of custom styling and brand assets from the start of the creation process. ## **Key Benefits** * **No technical skills required**: Apps are generated directly from text descriptions. * **Data integration**: Apps connect to workspace data and business tools such as CRM, calendar, Slack, or email. * **Built-in AI capabilities**: The system supports text generation, summarization, structured data extraction, and image analysis. * **Interactive preview**: Generated apps run immediately in an interactive preview environment. * **Iterative refinement**: Modifications are applied directly to the existing app without rebuilding it from scratch. * **Secure execution**: Apps run in isolated sandboxes with controlled data access. ## **When to Use** * **Dynamic or interactive outputs**: Best suited for dashboards, exploratory views, walkthroughs, and lightweight tools. * **Self-contained workspace tools**: Ideal for tools that live in the workspace and are shared with team members. * **Data and AI integration**: Best for combining workspace data with AI capabilities in a tailored interface. * **Custom-branded experiences**: Ideal when apps require specific brand styling, logos, or visual assets. ## **What Can Be Built** ![Image 4](https://mintcdn.com/nuwacomgmbh/xhfIQQsia-TNL1Zs/images/image-4.png?fit=max\&auto=format\&n=xhfIQQsia-TNL1Zs\&q=85\&s=27aa7ed259153e8ae6e99796628a904b) | Category | Examples | | - | - | | **Dashboards & reports** | KPI overviews, revenue/signups/conversion, order tracking | | **Tools & calculators** | Configurators, pricing calculators, internal helpers | | **Forms & intake** | Support intake, settings panels, filters, file uploads | | **Boards & maps** | Strategy, process, customer, or journey views | | **Infographics & explainers** | Architecture/system overviews, landing-style pages | | **Charts** | Focused single-dataset visualizations | | **Knowledge explorers** | Searchable interfaces over your knowledge base | | **Games & animations** | Small interactive experiences, New animations, and 3D scenes | ## **Branding and Styling** The system allows the integration of custom design standards and visual assets during the creation process. This supports several key use cases: * **Landing pages**: Creating custom pages with tailored visuals and consistent brand styling. * **Campaign pages**: Generating pages from a design brief with product screenshots and approved images. * **Sales tools**: Developing proposal configurators and company-branded microsites. * **Interactive dashboards**: Turning presentations, concept documents, or image sets into interactive dashboards. * **Internal tools**: Creating tools with correct logos, illustrations, and visual style from the start. * **Event and launch pages**: Building event pages, product launch pages, or executive demos with a strong design standard. ## What Apps Can Do Beyond appearance, apps perform real tasks. Everything below works without writing any code and is requested in chat. All access runs securely through the Nuwacom platform, so apps never handle credentials and can never reach the internet directly. ### Work with your data * **Read workspace data**: knowledge base files, projects, agents, and content serve as live data sources. * **Connect business tools**: live data from connected tools such as CRM, calendar, email, and Slack appears inside the app. * **Connect MCP servers directly**: an app binds straight to a tool on an MCP server as a live data source, with no agent in between. Depending on the tool, this can read data and also perform actions (read/write). * **Search the web**: current external information such as weather, news, prices, and look-ups is pulled in through the platform's built-in web search. ### Take action * **Send emails and messages**: apps trigger real actions in connected tools, for example sending an email (Outlook, Gmail) or posting a Slack message. These run under each person's own connected account and are available to people with **edit** access to the app. * **Create and update records**: any action a connected tool or MCP server offers can be wired in, for example creating a ticket, updating a deal, or adding a calendar event. ### Use AI inside the app * Generate and summarize text, extract structured data into a defined shape, and run a configured Nuwacom agent from within the app. * **Images:** generate, edit, and analyze images (vision). ### Save and export * **Remember app state**: lightweight state such as a selected tab, a draft, a preference, or a score is saved so it survives a reload. State is per-app and per-person, stored in the browser, and does not sync across devices. * **Save files to your knowledge base**: a file, image, or export such as a generated report in Markdown or JSON is saved into the workspace knowledge base, up to 25 MB per file. * **Export**: the whole app downloads as a codebase ZIP or exports to PDF. ### Look and feel * **Rich visuals**: charts and data visualizations, maps, 3D scenes, animations, and drag-and-drop interactions. * **Sounds**: apps play sounds and audio feedback, for example a timer chime or a click. Sounds are generated or embedded in the app. > [!NOTE] > > Apps do not run their own server or shared database. They save lightweight state in the browser and files to the knowledge base, but there is no always-on backend storing records for everyone. For shared, structured data, connect a tool or an MCP server as a data source instead. ## What Apps Can't do * **No direct internet access**: apps cannot call external websites or APIs directly and cannot embed third-party widgets. All external data goes through the platform, either web search or connected integrations. * **No shared or central database**: there is no always-on backend that stores records for all users. Connected tools or MCP servers cover shared, structured data. * **Browser-only state**: saved app state lives in one browser, per person. It is not shared between users and does not sync across devices, so it serves convenience rather than important records. * **Audio limits**: apps play generated or embedded sounds, but cannot stream audio from an external website and cannot use the microphone. * **No camera or microphone capture**: apps cannot record from a camera or microphone. * **Actions need edit access**: sending emails or messages and other write actions require edit access to the app. Read-only viewers can view data but cannot trigger actions. * **No credentials or model choice**: apps never handle API keys, and a specific AI model cannot be selected. The platform manages both. * **Login always required**: apps have no public or anonymous link. Every user must be a logged-in workspace member. * **Fair-use limits apply**: see **Usage limits (fair use)** below. ## Apps Overview Opening Apps from the sidebar shows the central overview of every app available in the workspace. The *+ New App* button in the top-right corner starts a new build. Each app summarizes and exposes its main controls: * **Initials badge**: a two-letter avatar in a colored circle, for example "CS" for Customer Success Cockpit. * **App name**: the bold title that identifies the app. * **Star icon**: marks the app as a favorite for quicker access. * **Description**: two to three lines of truncated text describing what the app does. * **Workspace label**: indicates that the app belongs to the workspace. * **Three-dot menu**: opens actions for the app, including Pin to sidebar, Edit, Rename, Duplicate, Share, and Delete (shown in red). ### Filtering, Searching, and Sorting The overview provides three controls to narrow down the grid. They combine, so a search term and a sort order apply within the selected tab. ### Tab filter Three tabs control which apps appear: * **All**: every app in the workspace, which is the default selection. * **My apps:** only apps created by the current user. * **Shared with me**: only apps that other users have shared with the current account. ### Search bar The search field filters the grid by app name as characters are entered. Search applies within the active tab, so a term entered under *Shared with me* searches only shared apps. ### Sort menu The *Recently Used* control on the far right opens the sort menu, which reorders the grid by one of the following criteria: * Recently Used * Most Used * Recently Created * Recently Updated * Alphabetically ## How It Works 1. **Open Apps and click +New App**: start a new build from the Apps overview. 2. **Describe**: for example, "Build a dashboard with our project numbers" or "A weather app for Berlin". 3. **Watch**: the app is generated and appears live in the preview on the right. 4. **Refine**: content, layout, colors, and features are adjusted through chat until the app is right. ## **The App Workspace** Opening an app shows a split view: the app canvas on the right and an optional chat sidebar on the left for AI-assisted editing. The canvas offers these views: | **View** | **What it shows** | | - | - | | **Preview** | The live, interactive app: click through it like a real app | | **Code** | The generated source files in a read-only editor with a file tree | | **Sources** | Files and content connected to the app as data | | **Integrations** | Actions from connected business tools (e.g. Outlook, Slack) | > [!NOTE] > > Read-only viewers see only the running Preview; all editing controls are hidden. ### **Preview** The app runs live and is fully interactive (forms, buttons, navigation). Internal links navigate within the app; external links open in a new browser tab. The app can be expanded to fullscreen. The preview refreshes automatically when data sources or versions change. ### **Code** A read-only view of the app's source, with a browsable file tree and syntax highlighting. From here, you can perform the following actions: * **Copy** Copy a file's contents directly to the clipboard. * **Download file** Download a single file from the codebase. * **Download codebase** Download the entire codebase as a ZIP archive. * **Reference in chat** Reference a specific file in chat to guide the AI when requesting targeted changes. ## **Connecting Your Data** Apps do not just show static content; they connect to live data, managed visually in the canvas: * **Sources**: attach files or content from the knowledge base as runtime data. After a source is added, an optional "smart apply" step describes how it should be used, and the AI wires it in. * **Integrations**: connect specific actions from business tools such as CRM, calendar, email, and chat. The action is selected, parameters are configured, a connection is chosen, and the setup is tested before saving. * **MCP servers**: bind directly to a tool on an MCP server as a live data source, without an agent in between. This reads data and, where the tool supports it, writes. > [!NOTE] > > Credentials are never entered directly within an app. The platform securely brokers all access. **In shared apps, each user connects and uses their own accounts; connections are never shared between users.** ## **AI Capabilities** Apps use AI directly, including: * Text generation and summarization. * Structured data extraction that returns results in a defined shape. * Running a configured agent from inside the app for multi-step or reasoning tasks. * Image generation, editing, and analysis (vision). AI actions are subject to the fair-use limit of 30 per minute (see Usage limits (fair use)). The model is not selected or paid per use; the platform picks an appropriate model. ## **Live and External Data** Apps can pull in current web information (weather, news, prices, look-ups) through the platform's built-in web search. All external data is routed securely through nuwacom; apps cannot call arbitrary external services directly. ## **Usage Limits (Fair Use)** To maintain platform stability and performance, each app is subject to built-in fair-use request limits. These limits prevent individual apps from overloading the system (for example, due to infinite loops) and are identical across all subscription plans and trials. * **Short-burst limit** — Apps can send a rapid burst of up to **20 requests** simultaneously, after which they continue at a steady rate of approximately **10 requests per second**. This is sufficient for standard navigation but may be briefly exceeded by data-heavy screens (e.g., a dashboard loading dozens of metrics at once), resulting in a temporary "rate limit exceeded" message. * **AI limit** — AI-driven actions (such as text generation, summarization, or image analysis) are capped at **30 requests per minute** to prevent runaway loops and unexpected costs. | Limit | Value | | - | - | | Short burst | up to 20 requests at once | | Steady rate after the burst | about 10 requests per second | | AI actions | 30 per minute | ## **Handling Rate Limit Messages** * **Reload the app** The limit clears within a second or two, so a reload usually resolves the issue. * **Reduce initial load** If the message persists, the app may be loading too much data at once. Instruct the AI in chat to load data in smaller steps, show fewer items initially, or load details only upon clicking. ## **Iterating and Editing** Editing is conversational. The AI makes precise, targeted edits rather than rebuilding the entire app. Users can choose between two build quality tiers: * **Standard** Faster and more cost-efficient. * **High** Higher quality for complex work. ## **Versions and History** Every change accumulates as a version snapshot. From the canvas, you can manage versions using these features: * **Open version history** Access the version history navigator from the canvas. * **Navigate versions** Step back and forward through previous version snapshots. * **Restore versions** Restore an earlier version as the active, live app. > [!NOTE] > > Historical versions are read-only snapshots; live data connections are inactive during preview. ## **Saved State** Apps have three ways to keep data, and the differences matter: * **App state (in the browser)**: lightweight state such as a selected tab, a draft, a preference, or a small score that survives a reload. It is per-app and per-person, stored in the browser. It is not shared with other users and does not sync across devices. * **Files in the knowledge base**: apps save a file, image, or export into the workspace knowledge base (up to 25 MB per file), where it becomes a normal knowledge base asset for reuse elsewhere. * **Live data from connected sources**: for shared, always-current, structured data, apps read from the knowledge base, connected tools, or MCP servers rather than storing it themselves. There is no built-in shared app database: apps do not run an always-on backend that stores records for everyone. When that is needed, connect a tool or MCP server as a data source. ## **Sharing and Permissions** The app and the chat are shared separately, each via its own Share button. Sharing one does not share the other. Apps are shared from the Share dialog, where you can perform these actions: * **Add individual users or groups** Grant access to specific team members or departments. * **Grant Read or Edit access** Configure permissions (new users default to Read access). * **Share with all workspace users** Make the app accessible to the entire workspace (subject to role permissions). * **Copy a link** Generate a 7-day invite link for logged-in workspace members. Sharing an app shares the finished app, not the building conversation. If the chat is shared, recipients can read the full message history in a read-only format. ### **Privacy Warning for Connected Accounts** > [!NOTE] > > Shared apps never share credentials or connections. When an app uses private connected accounts, each recipient uses their **own** connected account. The first time a viewer opens a shared app requiring an unlinked connection, a prompt appears to connect an account. If they do not connect, the affected part of the app remains unresolved. ### **Who Can Build vs. View** | **Capability** | **Editor** | **Read-only viewer** | | - | - | - | | Run the app (Preview) | Yes | Yes | | Preview / Code / Sources / Integrations toggle | Yes | Preview only | | Edit via AI chat | Yes | No | | Add/remove data sources and integrations | Yes | Preview only | | Download codebase | Yes | No | | Share | Yes | Hidden | | Delete | Yes | Hidden | ## **Accessing and Running Apps** * **Apps list** Find, search, open, rename, share, move, and delete apps from the Apps area in the sidebar. * **Dedicated app page** Each app has its own full-page view combining the live app and an editing chat sidebar. * **In chat** Apps appear as preview tiles; click to open the full canvas. Apps run inside nuwacom in a secure, sandboxed environment via a short-lived authenticated session. Access always requires a logged-in workspace member. ## **Export and Download** * **Download codebase** ZIP of all source files, available from the Code view. * **Download as PDF** Available from the app's "more" menu. ## **Security and Sandboxing** Apps are built to be safe even though they run real, interactive code: * **Isolated execution** Each app runs in a sandboxed frame on its own isolated address, separated from the main platform and other apps. * **No uncontrolled data egress** Apps cannot make direct calls to arbitrary external services, and other data-leak channels are locked down. * **Controlled, secure data access** Apps access workspace data, AI, and connected tools only through a controlled platform bridge. * **No credential exposure** Apps never see authentication tokens or credentials; the platform brokers every connection. * **Per-user connected accounts** Shared apps use each viewer's own connections. ## **Good to Know / Current Limitations** * **No element-level visual editing** Changes are made through chat rather than drag-and-drop or click-to-style controls. * **No manual preview reload button** The preview refreshes automatically when changes are applied. * **No public or anonymous URLs** Access always requires a logged-in workspace member. * **Title rename location** Renaming is performed from the Apps list, not the canvas header. * **Automatic version snapshots** Snapshots are created automatically from edits; there is no manual save button. * **Client-side execution** Apps run client-side only and cannot call external services directly. * **Fair-use request limits** Apps are subject to built-in request limits to maintain platform stability. ## **Example Prompts** * *“Create a KPI dashboard with revenue, new customers, and conversion.”* * *“Build a support intake form with file upload.”* * *“Show our projects as a board grouped by status.”* * *“Create an infographic of our system architecture.”* * *“Add a dark mode toggle to the header.”* ## **FAQ** > **Do viewers see my chat history?** > > It depends on **what** you share, because the app and the chat are shared separately: > > * **If you share the app** (Share button on the app): recipients get only the finished, running app. They do **not** see the chat you used to build it, and read-only viewers can't open the editing chat at all. > * **If you share the chat** (Share button on the conversation): recipients can open the conversation and read the **full message history**. It is read-only for them (they can't send messages), but they do see what was written. Only share the chat when you actually want to show how the app was built. > **Can I keep editing after I've shared an app?** > > Yes. Editing continues through chat, and people with access see the updated app. > **Can an app call external APIs or embed third-party widgets directly?** > > No. For security, apps cannot make direct external calls. Live or external data is provided through the platform (built-in web search) or through connected integrations. > **Do I need to manage API keys or credentials?** > > No. The platform securely brokers all data and tool access; apps never handle credentials. > **Can read-only users change the app?** > > No. Read-only users can run and interact with the app, but editing, data-source changes, downloading the code, sharing, and deleting are restricted to editors. > **Whose connected accounts does a shared app use?** > > Each person uses their own. When an app relies on an integration, recipients are prompted to connect or select their own account. > **I saw "rate limit exceeded" when opening an app, what does that mean?** > > It is a built-in fair-use limit on the number of requests an app can send in a short burst. It usually appears when a data-heavy app loads a lot of data at once. Reloading the app or instructing the AI to load data in smaller steps resolves the issue. # Introduction to Connectors Source: https://docs.nuwacom.ai/en/product/connectors ### Overview Connectors link nuwacom to the tools an organization already uses, making external data and actions available directly within Chat, Agents, Knowledge, Workflows, and Apps without switching between applications. They cover both Connectors and MCP servers, enabling anything from simple integrations to custom workflows that access relevant data and perform actions in connected systems. nuwacom offers two types of Connectors, Connectors and MCP Servers. Both are managed by admins under *Workspace settings > Connectors*. ### Connectors [Connectors](/en/connector-directory) are preconfigured integrations for common business tools. They cover typical use cases and are set up quickly without additional configuration. Once an App is connected, agents use it to retrieve information, analyze content, and perform actions in the connected tool. Available Connectors include Slack, Microsoft Teams, Microsoft Outlook, Gmail, Google Docs and more. ### MCP Servers [MCP servers](/en/mcp-directory) enable deeper and more flexible integrations beyond standard Connectors. MCP stands for Model Context Protocol and provides a standardized way to connect external systems, so agents not only access data but also execute more complex and granular actions across systems. MCP servers suit advanced use cases, custom integrations, and technical workflows, and are used primarily by administrators and advanced teams. A connection is created from a catalog of common MCP servers, or a custom server is added manually when needed. Available MCP servers include GitHub, Atlassian, Notion, Linear, Canva and more. ## Key Benefits * Checking a Google Calendar for availability and scheduling a meeting * Looking up a contact in a CRM to draft a follow-up email * Searching a Jira backlog or creating a new issue * Pulling a Confluence page into a conversation as reference * Sending a Slack or Microsoft Teams message to a team * Retrieving, analyzing, and saving files in Google Drive, OneDrive, or SharePoint nuwacom uses these Connectors to retrieve information, analyze content, and execute actions in connected tools, such as reading emails, fetching files, or managing calendar events. ## When to Use Connectors Connectors are the right choice when nuwacom needs to interact with tools a team already relies on. They retrieve data, search company sources, trigger actions, and connect agents and workflows to external systems. For most everyday tools, an App provides a fast, preconfigured path. For tools that require deeper access, custom logic, or self-hosted systems, an MCP server offers the necessary flexibility. ## Connector Governance Connectors are managed by admins under *Workspace settings > Connectors*. Admins can enable or disable each Connector individually, connect accounts, and control whether access remains private, is limited to selected team members, or is shared with the entire workspace. *** ## FAQ > **What are Connectors and what are they used for?** > > Connectors bring data from connected business tools directly into nuwacom, so files and information from tools like Google Drive or SharePoint become available as a knowledge source in chats. They also let agents run actions in those systems, from sending emails automatically to managing project documents. > **What happens to data when business tools are connected to nuwacom?** > > Connected tools follow the same strict data protection standards as all nuwacom services: no permanent storage, encryption in transit (TLS 1.2/1.3) and at rest (AES-256), full GDPR compliance, and hosting exclusively in ISO/IEC 27001- and SOC 2-certified data centers in Germany. > > Access remains private or is shared with team members or the whole workspace. Sharing a Connector grants recipients access to all available data of the connected account. > **What happens to emails when an email account is connected?** > > Agents access emails only on explicit instruction, for example, "search for emails from X" or "send an email to …". Email content is processed temporarily rather than stored permanently. > > All communication runs over encrypted connections (TLS 1.2/1.3), access is governed by role-based access control (RBAC) with full audit logging, and the connection can be revoked at any time via secure OAuth. > **Is data from connected tools used for AI training?** > > No. nuwacom does not use company data to train AI models and, as a GDPR-compliant data processor, processes data only according to the provided instructions and contractual agreements. > > Customer data is strictly separated by purpose, and all subprocessors, including AI providers, are bound by equivalent security standards and written contracts. > **Can the available Connectors in a workspace be controlled?** > > Yes. Administrators define which Connectors a team can use in workspace settings under *Connectors* via *Manage Connectors*, where each Connector can be enabled or disabled as needed. > > This keeps the workspace secure and compliant while maintaining control over external tool integrations. > **How is a Connector shared securely with a team?** > > Sharing starts in workspace settings under *Connectors*. Select the *lock icon* next to the Connector in the *Connected Accounts* list, choose the relevant team members or the entire workspace, and confirm with *Done*. > > The principle of least privilege applies: access should remain limited to people who need it, shared access should be reviewed regularly, and highly sensitive accounts such as executive email accounts should remain private. > **How is a Connector disconnected or access revoked?** > > A connection can be removed in workspace settings under *Connectors* by selecting the relevant Connector in the *Connected Accounts* list. > > nuwacom's OAuth permission can also be revoked directly in the connected service's settings, for example in Google or Microsoft. > > At contract end, all customer data is securely returned or deleted, with all copies and backups destroyed within 90 days. # Getting started Source: https://docs.nuwacom.ai/en/product/gettingstarted **Welcome to nuwacom! 🎉** \ You’ve just joined a platform designed to simplify how your team works with knowledge, documents, and AI. This guide walks you through your very first steps — from logging in, to setting up your brand, inviting your team, and trying your first real use cases. By the end, you’ll be ready to collaborate and create confidently with AI.  ### What is nuwacom? Imagine one workspace where all your knowledge, tools, and conversations are connected. That’s nuwacom. Instead of jumping between SharePoint for documents, Slack for communication, and Google Drive for files, you now have a single place where:  * Your knowledge sources are connected.  * AI helps you research, create, and organize content.  * Your team collaborates in real time.  It’s more than a tool — it’s your **AI-first operating system for work.**  *** ## **1. Use the Chat & First Prompt** Meet the heart of nuwacom — the Chat.  ![Chaten](https://mintcdn.com/nuwacomgmbh/dP_XChvbzqsyD4Jb/images/chaten-1.png?fit=max\&auto=format\&n=dP_XChvbzqsyD4Jb\&q=85\&s=e29b44396cc1bf01a5c634abf3531a7a) Think of chat as a supercharged workspace where you can:  * Ask questions based on your documents.  * Search the web in real time.  * Use saved prompts for repeatable tasks.  **Try this:**  * Open a new Chat.  * Type: *“Summarize last quarter’s sales deck into three takeaways for leadership.”*  * Upload the Sales Deck document and watch the AI highlight the key insights for you.  * Use a follow-up prompt (e.g., “Create an executive summary”) to structure the result.  👉 You’ve just run your first prompt — and seen how AI + your knowledge base work together.  *** ## **2. Try Some Use Cases** Now, you know the basics. Test nuwacom with real scenarios your team faces every day – for example, by working on typical tasks like creating documents, drafting emails, building presentations, or conducting research. ![Research](https://mintcdn.com/nuwacomgmbh/dP_XChvbzqsyD4Jb/images/research-2.png?fit=max\&auto=format\&n=dP_XChvbzqsyD4Jb\&q=85\&s=055fd8f301427843dc07e10c1d2dcdbf) ### **Do Research**  * Upload a complex document and let the AI summarize it in seconds  * Quickly understand long reports or presentations  * Ask follow-up questions in chat to explore specific details or insights.  ### **Write Content**  * Open the chat and type *“Draft a LinkedIn post.***”** Add context — such as the topic, target audience, and communication goal — either in your message or by attaching a document. The AI then uses this information to generate a precise, well-aligned draft.  * Apply your brand voice to keep it consistent.  * Edit in AI editor. * Share with colleagues to edit with you in real time.  ### **Upload a Document & Analyse**  * Upload a PDF, PowerPoint, or Word doc.  * Ask: *“Highlight the risks”* or *“Explain this in plain English.”*  * The uploaded file will be stored in Knowledge Base>Private files for future search.  ### **Create an Image**  * Open chat and click *Create Image* to create blog visuals, campaign graphics, or diagrams.  * Add brand styles like logos or colors for consistency.   * Ask: *Create a social media image for our new electric bike launch. Use our brand colors and show a modern city background with a rider on an e-bike.*  👉 The more you experiment here, the quicker you’ll see where nuwacom fits into your workflows.  *** ## **3. Create & Manage Prompts** Soon you’ll notice you’re asking AI the same types of questions. Prompts let you save these instructions so you don’t start from scratch each time.  **What to do:**  * Type @ to open Prompts.  * Browse pre-built prompts (e.g., “Create a blog post”, “Create a LinkedIn post”) — available with the Teams plan. * Create your own and save it for reuse.  * Share it with your workspace so everyone benefits.  👉 Example: Instead of typing *“Summarize customer feedback into positive and negative points”* every time, save it once as a prompt called *Customer Feedback Summary.*   *** ## **4. Create & Manage a Project** Projects help you keep everything related to a goal in one place. Imagine planning a product launch: you’ll have research, content drafts, meeting notes, and images all connected.  ![Documents 20250916 081204 Jp](https://mintcdn.com/nuwacomgmbh/iV4eMOjbLDKUIZpe/images/Documents-20250916-081204.jpg?fit=max\&auto=format\&n=iV4eMOjbLDKUIZpe\&q=85\&s=561b53d0ed1d4d0fa8d81b8d78ebf1fb) **What to do:**  * Go to Projects and create a new one.  * Add files to your project — it will be used as context for all chats and documents.  * Add AI guardrails (e. g. communication style, brand rules, wordings, facts) which will be considered in the whole project  * Use it to generate multiple outputs: blog posts, newsletters, social posts — all consistently based on the same knowledge.  * Assign responsibilities to teammates and track progress.  👉 Projects make collaboration simple. Everyone sees the same context, so there’s less back-and-forth.  *** ## **5. Record Meetings & Automate Follow-Ups** Capture meeting content in one place with or without a bot.Record meetings directly, use the Meeting Bot, or add manual notes. Transcripts and summaries are generated automatically and combined with your own input. ![Meetingnotes](https://mintcdn.com/nuwacomgmbh/dP_XChvbzqsyD4Jb/images/meetingnotes-2.png?fit=max\&auto=format\&n=dP_XChvbzqsyD4Jb\&q=85\&s=e1fb52b8703fe2096c1efd1edc723a10) **What to do:**  * Record your meeting through nuwacom.  * AI generates a **summary**, **key decisions**, and **action items**.  * Share them with participants or link them directly to an ongoing project.  👉 No more “What did we agree on?” moments — everything is captured and actionable. *** ## **6. Share & Collaborate** Finally, nuwacom is about working together. You can share documents, prompts, or entire projects with teammates.  ![Sharechat.png](https://mintcdn.com/nuwacomgmbh/GeTLputeVSiVLUhO/images/Sharechat.png?fit=max\&auto=format\&n=GeTLputeVSiVLUhO\&q=85\&s=567385179edde24fc61ba86cd1da2e31) **What to do:**  * Click *Share* on any item.  * Collaborate in real time — see updates instantly, just like in Google Docs, but with AI helping in the background.  👉 This makes nuwacom your team’s single source of truth for AI-driven work.  *** ## Need help? We’re here for you. You can reach our nuwacom-Bot or our team anytime right from your app. Just open the support messenger at the bottom left. Ask a question or browse our Help Center for quick answers and step-by-step guides. ![Supportmessenger.png](https://mintcdn.com/nuwacomgmbh/GeTLputeVSiVLUhO/images/Supportmessenger.png?fit=max\&auto=format\&n=GeTLputeVSiVLUhO\&q=85\&s=1640bad1c813747bba4c3370251da6de) *** ## What’s Next After learning the basics, you’re ready to explore the [**advanced features**](https://docs.nuwacom.ai/en/product/advanced) — from building custom agents and managing your knowledge base to setting up connectors and automating workflows. # Knowledge Source: https://docs.nuwacom.ai/en/product/knowledge ## Overview Files, notes, and connected sources — all in one place. Knowledge is the central store that supplies chats, agents, and projects with up-to-date information, keeping every relevant document structured and automatically synced. Everything a team relies on comes together in one searchable place, so content created natively in nuwacom and content uploaded or synced from external tools such as SharePoint or Google Drive can be found, organized, and reused from a single interface. > [!NOTE] > > Data protection comes first. All information is handled according to the highest security standards. Your sensitive data stays protected and is only used for the purposes you define. ## Key Benefits * All sources, including Data Connectors such as SharePoint and synced spaces and folders from Confluence or Google Drive, appear directly in the sidebar. * New content is created in place, generating documents, slides, notes, emails, images, and videos directly within the Knowledge. * Several files can be selected at once to start a chat, move, share, or download them together, and a chat can also be started directly from a single file. * Content is found fast through full-text search across the entire Knowledge, combined with filters by type, source, date, person, or project. * Team collaboration is supported, allowing root- and subfolders to be shared with individuals or the entire team while access rights are managed precisely. ## When to Use Knowledge Knowledge is used whenever a company's collective information — connected sources and stored content — needs to come together in one place to support work with AI. Documents, notes, files, and connected sources provide the information that chats, agents, and projects draw on to answer questions, generate content, analyze data, and support workflows. *** ## Knowledge Overview In Knowledge, content is organized across several key areas: * *My Drive* contains personal content * *Shared with me* collects files and folders that others have shared. * The table view lists the contents of the current location in a structured, sortable format. * *Searchbar* searches across the entire knowledge ## The Table View Files, notes and connected sources are presented in a table. Each column supports sorting, and a checkbox on each row enables multi-select for batch actions. | Column | What it shows | | :- | :- | | Name | The file, folder, or note title, with a type icon | | Shared | The people the item is shared with, shown as avatars | | Created by | The person who created the item | | Modified | The date the item was last changed | | Type | The resource format, such as Folder, PDF, Image, or Note (optional) | | Source | The origin of the item, such as nuwacom or a synced connection (optional) | | Size | The file size (optional) | | Created date | The date the item was originally created (optional) | The table can be sorted by clicking a column header to toggle between ascending and descending order. ## Filtering Two controls narrow the current view; both can combine with each other and with search to refine results results down to a precise set. The *People* button filters content by person. A *Search people* field locates someone quickly, and three scopes refine the results: *All* returns every item associated with that person, *Creator* limits results to items they created, and *Shared with* limits results to items shared with them. The *Filter* button narrows the view by attribute: * *Type* filters by resource format, such as folders, PDFs, images, or notes. * *Source* filters by origin, such as nuwacom or a specific synced connection. * *Modified* filters by time frame, with options including *Any date*, *Last 24 hours*, *Last 7 days*, *Last 30 days*, and *Last year*. * *Created* filters by the date the item was created, including *Any date*, *Last 24 hours*, *Last 7 days*, *Last 30 days*, and *Last year*. * *Project* filters by an associated project, with a *Search projects* field to locate one quickly. ## Add files The *+Add* button brings existing content into the Knowledge. Its menu covers: * *New folder* creates an empty folder to organize content. * *Upload folder* imports an entire folder and its contents. * *Upload files* imports individual files from your device. * *Synced connections* links external services such as SharePoint and Google Drive, so their files stay available and searchable inside the workspace. > [!NOTE] > > **Important:** > > * All content is **private by default** and only visible to you > * Items created in nuwacom (Notes, Documents, Slides etc.) are **automatically saved here** > * You decide at any time if and with whom you share content ### Supported File Formats nuwacom supports a range of file formats for uploading and processing content. > **View all supported file formats** > > **Plain text & markup** > > * `application/json` > * `text/plain` > * `text/markdown` > * `text/html` > * `application/xml` > * `text/xml` > * `text/x-rst` > * `application/rtf` > * `text/rtf` > > **Emails & messages** > > * `application/vnd.ms-outlook` > * `message/rfc822` > > **Images** > > * `image/jpeg` > * `image/png` > * `image/webp` > * `image/gif` (non-animated) > > **Documents (Office & standard)** > > * `application/msword` > * `application/vnd.openxmlformats-officedocument.wordprocessingml.document` > * `application/vnd.oasis.opendocument.text` > * `application/epub+zip` > > **Tables & spreadsheets** > > * `text/csv` > * `application/vnd.ms-excel` > * `application/vnd.openxmlformats-officedocument.spreadsheetml.sheet` > > **Presentations** > > * `application/vnd.ms-powerpoint` > * `application/vnd.openxmlformats-officedocument.presentationml.presentation` > > **PDF** > > * `application/pdf` > > **Audio & Video** > > * `audio/mpeg` > * `audio/wav` > * `audio/ogg` > * `audio/flac` > * `audio/aac` > * `audio/aiff` > * `audio/mp4` > * `video/mp4` > * `video/webm` > > **Google Drive** > > * `application/vnd.google-apps.document` > * `application/vnd.google-apps.spreadsheet` > * `application/vnd.google-apps.presentation` > > **nuwacom documents** > > * nuwacom documents > * nuwacom meeting notes > * nuwacom emails > * nuwacom slides ### File upload and storage limits File upload and storage limits apply per organization and depend on the subscription plan and deployment type. | Plan | Deployment type | Max. file size | Maximum number of files | | :- | :- | :- | :- | | Free | Multi-Tenant (app.nuwacom.ai and app-de.nuwacom.ai) | 100 MB | 1,000 | | Business | Multi-Tenant | 100 MB | 15,000 | | Enterprise | Multi-Tenant | 100 MB | 200,000 | | Enterprise | Dedicated Deployment | 100 MB | Unlimited | ## Create content The *+Create* button generates new content directly inside the Knowledge. The new item is created in the current location and appears in the table once saved. Available formats include: * Document writes structured content such as concepts or reports. * Slides generates presentation slides for meetings or pitches. * Notes captures ideas or meeting notes. * Emails drafts professional emails. * Images creates visual content from a description. * Videos creates videos based on text descriptions. *** ## Work With Selected Files Selecting a file in the Knowledge view opens a panel with a preview and two direct actions: *Start chat* and *Edit*. Both keep the selected file as the working context, so a file can move into a conversation or into its editor without being reattached. ### Steps: Start a Chat from a File 1. Open the Knowledge and locate the file through the folder structure or the *Search knowledge* field. 2. Select the file to open the preview. 3. Choose *Start chat* to open a new conversation with the file set as attached file. 4. Describe the task in the input field, then send the request. ### Steps: Edit a File 1. Select the file in the Knowledge to open the preview. 2. Choose *Edit* to open the file in its editor. 3. Apply changes using the editor tools. The editor that opens depends on the file type, with Slides opening in the slide editor and other file types opening in their respective editors. *** ## How do I organize files effectively? The Knowledge Base is the shared foundation for collaboration between your team and the AI. Its content is used as context so the AI can generate answers and content based on your data. Each team organizes its own knowledge area and takes responsibility for its content and structure. **Examples:** * The marketing team maintains a “Marketing” folder with brand guidelines, campaign materials, product copy, and social media templates. * The sales team organizes a “Sales” folder with presentations, proposal templates, pitch guides, and pricing sheets. * The legal or compliance team manages a “Legal & Policies” folder with contracts, data protection policies, approvals, templates, and approved wording. **💡 Tips for data hygiene** * Start lean with a clear and intuitive folder structure. * Structure folders by department, team, or use case. * Only use relevant and up-to-date content and avoid outdated files. *** ## How do I add files? Uploaded files and newly created folders are private by default and only visible to you. You decide when and with whom to share your content. Create **folders** to upload important documents and structure them clearly by department, team, or project. Each team—whether marketing, sales, product, or legal—can create its own folder with subfolders: * **Clear ownership:** Each team manages its area independently * **Faster onboarding:** New team members instantly find structured knowledge * **Scalability:** New teams and folders can be added anytime **Example:**\ Create a marketing folder and structure it with subfolders like *Brand Guidelines*, *Campaigns*, *Product Copy*, or *Social Media*. Upload all relevant documents so the AI can generate on-brand, consistent content and answer questions based on up-to-date information. **How it works:** 1. Navigate to the Knowledge Base. 2. Create a new folder and give it a clear name, e.g. “Marketing”. 3. Upload the relevant files your team and AI should use as context. You have multiple options: * **Sync external connectors:** Connect files from OneDrive, Google Drive, or SharePoint. * **Upload files:** Upload individual files such as PDFs, presentations, Word documents, or spreadsheets. * **Upload folders:** Upload entire folder structures at once. Subfolders and content are included automatically. * **Drag & drop:** Drag files or folders directly into the Knowledge Base. * **Create document:** Create new content directly in nuwacom, such as notes, briefings, or guidelines. 4. Invite your team to use this knowledge together. 5. Add the folder to chats, agents, and projects. When you update it, changes are automatically reflected everywhere. > [!TIP] > > **Tip:** Add entire folders or individual files as knowledge when configuring agents so they always work with the right context. [Learn more.](/product/agents#knowledge) *** ## How do I share files with my team? Uploaded files and newly created folders are private by default and only visible to you. You decide when and with whom to share your content. ### **How sharing works** 1. Open the folder and click Share. 2. Select one or more people or share the folder with the entire workspace. 3. Set permissions for each person (or the workspace). ### **Permissions** * **View**: View files but not edit them. * **Edit**: View, add, modify, or delete files. * **Remove access**: Revoke access at any time. **Example**\ A “Policies & Templates” folder starts as private and is then shared selectively—for example, with view access for marketing and edit access for the legal team. > **Important: Permission inheritance** > > * *Root folders and subfolders can be shared* > * *When you share a root folder, others get access to all included subfolders and files* > > **Example**\ > If you share the root folder “Policies & Templates” with someone, they automatically get access to all included subfolders and files. The assigned permissions (view or edit) always apply to the entire root folder. *** ## Sync external connectors Connect files from OneDrive, Google Drive, and SharePoint directly to your knowledge base. This ensures you always work with up-to-date content in nuwacom, without manually uploading or maintaining files. The AI extracts the content and uses it as the foundation for precise answers in chats, agents, and projects. **Your benefits** * **Automatic synchronization:** Changes in external data sources are reflected in your knowledge base in real time. * **Preserved structure:** Folder structures remain intact. * **No manual updates:** The AI always uses the latest content automatically. * **Single source of truth:** No version conflicts, no duplicates. > [!NOTE] > > Note: Only admins and super admins can sync external connectors with the knowledge base. ### **How it works** 1. Navigate to the knowledge base. 2. Select a section and click *Add*. 3. Choose the connector you want, for example SharePoint. 4. Connect your account if you haven’t already. 5. Select folders or individual files with one click and add them to the folder. **Result:** Selected content is connected within seconds, stays automatically synchronized, and is initially only visible to you. The AI extracts the content and uses it as a knowledge foundation to deliver precise, context-aware answers in chats, agents, and projects. ### **FAQ** > **Who can set up external connectors?** > > Only admins and super admins can sync external connectors with the knowledge base. > **I don’t see external connectors. What can I do?** > > Connectors must be enabled in your workspace settings. Navigate to Workspace Settings > Connectors > Available Connectors to enable the connectors you need. > **Which connectors are supported?** > > Currently, you can connect Sharepoint, OneDrive, Google Drive, and Confluence. Contact our team if you want to connect additional data sources or learn more about custom integrations. > **In which direction does the synchronization work?** > > Synchronization is one-way. Changes in the external source are automatically updated in nuwacom. However, changes made in nuwacom are not written back to the external system. > **What happens when new files are added in the connector?** > > New files are automatically synchronized. This process may take a few minutes. > **Can synchronized files be deleted in nuwacom?** > > No. Deleting a file in nuwacom only removes it from the nuwacom folder. The original file in the external data source remains unchanged. > **How can I identify automatically synchronized files?** > > Files with automatic synchronization are marked with a small cloud icon next to the document name. This allows you to quickly see which files are synced automatically. > **Are the original files stored in nuwacom?** > > No. Only the content is extracted. The original files are not stored in nuwacom. > **Who can access synchronized files?** > > Files you sync in “My Files” remain private and are only visible to you. Files in a new knowledge folder are also private by default. They only become accessible to others once you actively share the knowledge folder with users, user groups, or the entire workspace. > **How can I share synchronized files with others?** > > Files synced from external connectors are initially only visible to you. To share them, you have two options: > > **1. Upload files directly into a shared knowledge folder**\ > Upload files directly into a knowledge folder that is already shared with others. > > **2. Share files afterwards**\ > Move files via drag & drop into a shared knowledge folder. The files will then automatically be visible to everyone with the appropriate permissions. > **Are permissions from external data sources applied?** > > No. Permissions from SharePoint, OneDrive, or Google Drive are not transferred to nuwacom. Access is determined by the knowledge folder where the synced files are stored. > **What happens if a folder is shared with users who don’t have access in the connector?** > > If a knowledge folder is shared with user A, even though they don’t have access in the connector, user A will still gain access to the synced content within nuwacom. Permissions in the source system remain unaffected. *** ## Web crawler With the integrated web crawler, you can automatically add content from publicly accessible websites to your knowledge base. > [!NOTE] > > Contact our team to learn more about custom integrations and to configure the web crawler for your specific requirements. **How it works** * **Continuous analysis:** The web crawler regularly checks defined websites * **Real-time detection:** Changes are automatically detected and synchronized * **Dynamic content:** Even changing websites stay up to date **Use cases** * Company websites * Product documentation * Public knowledge bases * Industry news and updates *** ## **FAQ: Frequently asked questions** > **Who has access to my storage?** > > Only you. This area is private and cannot be shared. You can share folders with specific individuals or the entire workspace. For each person, you can define whether they can view or edit the folder. Access can be changed or removed at any time. > **How should I structure my knowledge base effectively?** > > Structure your knowledge base to reflect your team’s daily work. Store relevant and up-to-date content centrally and organize it clearly, for example by departments, teams, or use cases. This ensures both employees and AI work with the same reliable knowledge base and produce consistent results. Start lean with an intuitive folder structure and avoid outdated or duplicate content. > **Can I link entire folders to chats, agents, and projects?** > > Yes, you can use folders as context for chats, agents, and projects. Update the folder once in the knowledge base, and changes will automatically apply everywhere. > **Who can upload and manage files?** > > Editors, admins, and super admins can upload files and share them with others, assigning view or edit permissions per user. > > Readers only have access to shared files with view permissions. They cannot upload or manage files. > **How do I share files with other users?** > > Files you upload are initially only visible to you. To share them, click on the root folder > Share and assign the necessary permissions. Subfolders cannot be shared. # Notetaker Source: https://docs.nuwacom.ai/en/product/notetaker ## **Overview** ![Notetakeren](https://mintcdn.com/nuwacomgmbh/BQOzyrXGKSQ4b5fb/images/notetakeren.png?fit=max\&auto=format\&n=BQOzyrXGKSQ4b5fb\&q=85\&s=0e20687022ed47447596e6939dc13774) The Notetaker automatically joins your meetings, records and transcribes them, and generates AI-powered summaries. Summaries are stored as searchable content in your workspace and can be referenced directly in AI chat - making meeting knowledge accessible long after the call has ended. The Notetaker connects to your calendar to detect upcoming meetings and join them without manual intervention. ## **Key Benefits** * Decisions, action items, and owners are captured automatically. No manual note-taking required. * Meeting summaries are indexed and searchable alongside all other workspace content. * Team members who did not attend can catch up from the structured summary without watching a recording. * Meeting knowledge feeds directly into AI chat, agent contexts, and projects. ## **When to Use Notetaker** The Notetaker is relevant whenever consistent, structured documentation of meetings matters: * Internal team meetings where decisions and follow-ups need to be tracked. * Customer and sales calls where commitments and next steps must be captured and searched later. * Recurring stand-ups or weekly team meetings where a consistent summary format is required. * Any meeting where not all stakeholders can attend live. ## **What You Can Do With It** * Review and edit the structured summary under *Recent notes* after each call. * Create and send follow-up emails automatically based on the summary. * Generate action items and transfer them to your calendar. * Use meeting notes as a knowledge source in Chat or Projects. * Share meeting notes with colleagues for transparent collaboration. * Regenerate the summary using a different template at any time. *** ## **Recent Notes** The *Recent Notes* section provides an overview of the notes available in the workspace. It includes notes generated by the Notetaker as well as notes created manually. ### **Find and filter notes** Three controls are available to narrow the list: **People** — Opens a searchable list of all workspace members. Selecting a person filters the notes list to show notes where that person is either the creator or a shared recipient. The "All," "Creator," and "Shared with" tabs narrow the scope further. **Filter** — Opens a filter panel with three categories: | Filter | Options | | :- | :- | | Type | Note, Folder | | Source | Google Calendar, Outlook, Zoom, Webex, nuwacom | | Modified | Today, Last 7 days, Last 30 days, Last 90 days | | Project | Search and select from available workspace projects | Each category expands into its own sub-panel and can be combined with the others. ### **Table columns** | Column | Description | | :- | :- | | Name | The title of the note. | | Source | The origin of the note — meeting service, manual entry, or integration. | | Shared | The people with whom the note is shared. | | Created by | The person who created the note. | | Modified | The date or relative time of the last change. | | Created | Note creation date | The table can be sorted by clicking a column header to toggle between ascending and descending order. ### **Use bulk actions** Select the checkbox next to one or more notes to open the available bulk actions: | Action | Description | | :- | :- | | Start new chat | Start an AI chat using the selected notes as context. | | Move | Move the selected notes to a project or knowledge base. | | Delete | Delete the selected notes. | The bulk action menu appears after selecting at least one note. *** # **Add note** The Notetaker provides a dedicated editor for capturing, formatting, and managing notes in nuwacom. Notes can be written manually, generated from a recording, or created from a connected calendar. Notes can also be shared, exported, or used to start a new chat. 1. Select *+ Add*. 2. Select *Add note*. 3. Enter the note content in the editor. ### **Note editor** The note editor provides a workspace for writing, formatting, and managing notes. 1. Enter a title in the *New Note* field. 2. Add content in the editor. 3. Use the formatting toolbar to format the content. The formatting toolbar includes the following options: * Select a text style from the *Text* menu. * Apply *cursive*, underline, or strikethrough formatting. * Highlight text. * Create bulleted or numbered lists. * Add links. * Insert tables. * Copy note content. ### **Record note** A note can be generated from a live recording. The spoken content is captured in the *Transcript* panel and converted into an AI Note. 1. Start the recording and speak the content. 2. Pause or resume the recording when needed. 3. Select *Generate note* to convert the transcript into a structured AI Note. The transcript can be copied, downloaded, or closed. AI Notes can also be created from a connected calendar. ### **Share note** Select *Share* to control who can access the note. Sharing gives others access to the same note, so collaborators can work with one shared version instead of separate copies. ### **Manage note** The note header provides the following actions: * *Share*: Share the note. * *Continue*: Start a new chat based on the note content. * *More options*: Open additional note actions. * *Close*: Exit the note editor. The *More options* menu includes the following actions: * Open the note in Google Docs. * Open the note in Word. * Move the note to a project. * Download the note as a DOCX file. * Download the note as a PDF file. * Delete the note. *** ## **How to connect a calendar** The following steps connect a new calendar so the Notetaker can capture meetings automatically. The flow applies to Google Calendar and Microsoft Outlook. > [!NOTE] > > Please ensure that all participants have given their consent before meeting notes are recorded automatically. 1. In the Notetaker, open the "Connect additional calendar" area. 2. Select the "+ Add calendar" button. 3. Choose the provider: Google Calendar or Microsoft Outlook. 4. In the provider's sign-in window, select the account to connect. If the account is not listed, add it via "Use another account." 5. Confirm the sign-in and review the requested permissions. Calendar permissions are required for automatic joining, including "See and edit events on all your calendars." 6. Confirm access with "Allow." 7. In the "Get insights from your meetings" dialog, select the appropriate join rule. 8. Apply the selection with "Save." "Cancel" closes the connection without changes. After saving, the Notetaker syncs the connected calendar's events. Upcoming meetings appear in the "Upcoming meetings" panel, and the toggle shows which meetings have automatic joining enabled. *** ## **Add Notetaker via meeting link** Alternatively, the Notetaker can be invited to any ongoing or upcoming meeting. Meeting links from Microsoft Teams, Google Meet, Zoom, and Webex are supported. > [!NOTE] > > Please ensure that all participants have given their consent before meeting notes are recorded automatically. ### **How it works** 1. Open the Notetaker. 2. Select *Join Meeting* 3. Paste the meeting link 4. Enter a meeting name (optional) 5. Set the date and time 6. Select "Add Notetaker" to confirm. The Notetaker joins the meeting as a guest at the scheduled time and must be admitted by a participant. Once admitted, it captures the conversation and automatically generates structured notes and a transcript. Please ensure that all participants have given their consent before meeting notes are recorded automatically. *** ## **Join rules** When connecting a calendar and in the calendar settings, the option "The Meeting Notetaker joins:" defines which meetings are recorded. Six rules are available: * **Join all meetings**: The Notetaker joins all meetings in the calendar. * **Join all accepted meetings**: The Notetaker joins all meetings you have accepted. * **Join meetings I organize**: The Notetaker joins all meetings you host. * **Join internal meetings**: The Notetaker joins all meetings held internally within your organization. * **Join external meetings**: The Notetaker joins all meetings that are external to your organization. * **Join when invited**: The Notetaker joins only when it is added manually. The general rule sets the default, while the per-meeting toggle overrides it for individual meetings. Settings can be configured per calendar provider (Google or Outlook). To disconnect a calendar, open Upcoming meetings, select the three-dot menu, and choose Remove calendar. *** ## **Supported Platforms** | Platform | Via | | :- | :- | | Microsoft Teams | Outlook Calendar sync & Meeting Link | | Google Meet | Google Calendar sync & Meeting Link | | Zoom | Meeting Link | | Webex | Meeting Link | *** ## **Bot Commands During Meetings** Control the Notetaker via chat commands during the meeting: | Command | Action | | :- | :- | | `/nuwa help` | Show available commands | | `/nuwa pause` | Pause recording (bot stays in the call) | | `/nuwa resume` | Resume recording | | `/nuwa leave` | Remove the bot from the meeting | *** ## **Meeting Summary Templates** nuwacom includes pre-built templates to structure summaries for different meeting types: * **Default Template:** General-purpose summary. * **1:1 Meeting Summary:** Focused on individual check-ins. * **Customer Discovery Summary:** Captures customer insights and feedback. * **Daily Stand-Up Summary:** Status updates, blockers, and next steps. * **Weekly Team Meeting Summary:** Topics, decisions, and action items. * **Brainstorm / Strategy Session Summary:** Ideas, themes, and conclusions. Custom templates can be created to match specific team needs. Templates can be set as default per user or per space, and can be switched after a meeting to regenerate the summary in a different format. *** ## **Summary Content** Each generated summary includes: * **Meeting Overview:** High-level recap. * **Participants:** Who attended, with roles where identifiable. * **Discussion Topics:** Organized by theme or domain. * **Decisions:** What was agreed upon. * **Action Items:** With owners and deadlines. The summary language automatically matches the language spoken in the meeting. To view the transcript, open the relevant meeting and select the *Transcript* tab. *** ## **Security and Data Privacy** * The Notetaker announces itself upon joining and informs all participants that the session is being recorded and summarized. * No audio or video recordings are stored. Only the text transcript and the generated summary are retained. * All meeting data is processed and stored within the EU. * Calendar sync, transcription, and summary generation run exclusively on EU-based infrastructure. * Meeting data is scoped to the workspace it belongs to. Only workspace members with the appropriate permissions can access summaries. * Role-based access control applies: editors and above can manage meetings; readers can view summaries. * Transcript data sent for summary generation is not used for model training by the AI provider. *** ## **FAQ** > **Can multiple people from the same workspace use the Notetaker in the same meeting?** > > Yes, but each person's Notetaker bot will join the call separately, resulting in multiple bots. To avoid this, coordinate who enables the Notetaker per meeting, or use the "Manual only" attendance setting and designate one person to activate it. > **What happens if my meeting does not have a link in the calendar event?** > > The Notetaker requires a meeting URL to join. Calendar events without a meeting link are skipped automatically. > **Can the summary be regenerated after a meeting?** > > Yes. The summary can be regenerated at any time using a different template, without re-running the meeting. > **Is the Notetaker limited in how many meetings it can process?** > > The Free Trial covers 5 meetings. Team and Enterprise plans include unlimited meetings when the meeting feature is enabled. > **Can the transcript be downloaded?** > > Yes. The transcript can be copied or downloaded from the **Transcript** panel. # Projects Source: https://docs.nuwacom.ai/en/product/projects Organize your documents in projects, create multiple documents, and collaborate with your team. Use projects to efficiently create and manage multiple pieces of content. Store central knowledge that serves as the foundation for all content created within the project, and edit your content with the help of AI. \ \ Add AI guardrails and rules. This way, you can produce brand-compliant and consistent content in no time. Of course, you can also share your content with individual team members or the entire workspace to collaborate.\ \ **Concrete examples of projects:** 1. **Marketing Campaign Launch** * A marketing team is preparing a new product launch campaign. * Inside the project, they create the campaign brief as a document, chat about design ideas, and attach brand guidelines. * They use prompts like “Draft LinkedIn Ad Copy” or call the *Content Agent* with “@AdWriter” to generate variations. * Everyone in marketing, design, and comms has access – so no information gets lost in emails or scattered docs. 2. **Client Proposal Preparation** * A sales team is working on a strategic proposal for a key client. * They gather past case studies, reference pricing sheets, and legal templates inside one project. * During chats, they can tag the *Proposal Agent* to draft sections or use autosuggested prompts to polish wording. * The project is shared with sales, legal, and finance – ensuring all stakeholders collaborate in one place. 3. **Product Development Sprint** * A product team sets up a project for the upcoming sprint. * User stories, specs, and research docs live in one place. * In chat, they call “@TechAgent” to generate code snippets or “@ResearchAgent” to summarize competitor benchmarks. * Designers, developers, and PMs all collaborate seamlessly without jumping across tools. 4. **HR Onboarding Program** * HR creates a project for new hires starting in October. * It contains onboarding documents, training checklists, and company guidelines. * The HR assistant agent helps generate personalized onboarding emails and FAQs. * Managers and HR share access so every department can align on a smooth onboarding experience. 5. **Compliance & Audit Preparation** * The compliance team prepares for an external audit. * All policies, contracts, and evidence docs are organized inside the audit project. * Prompts like “Summarize GDPR checklist” or “@PolicyAgent” ensure the latest legal text is applied. * The workspace is shared with compliance, legal, and external auditors – guaranteeing transparency and one single source of truth. *** ## **How do I create a project?** ### **1. Create a project** Simply click the 'Create Project' entry in the navigation. Enter your Project name in the dialog displayed and your project is created. ![Create\_Project.png](https://mintcdn.com/nuwacomgmbh/Xty9L1nK6TirbJ4W/CreateProject.png?fit=max\&auto=format\&n=Xty9L1nK6TirbJ4W\&q=85\&s=7be752e44498bfe449686b1c159c7ff0) **✔️ Add AI guardrails which will be considered in the whole project** Enter general guidelines for the project (e. g. communication style, brand rules, wordings, facts). This information will be used generating answers. **✔️Add files that are used as context within the project** The added files will be used as context information for all the chats and documents created in the project. * **Upload a new file:** Upload relevant files directly from your computer to nuwacom. * **Add from knowledge base:** Select documents from the knowledge base. ### **2. Create documents** Simply start chatting while the "Canvas" toggle is on. * Add predefined prompts and agents by clicking the plus icon in the chat and selecting "Run prompts and agents" * Add additional knowledge or files as context for this specific chat * Set your predefined brand voice via the "Select language style" icon *** ### **3. Optimize text in the AI editor** Once the document is generated, you can review and optimize it in the text editor using AI tools. Highlight text lines and apply AI adjustments, such as: * **Spelling and grammar:** Fix text errors. * **Tone of voice:** Adjust the text’s tone. * **Text simplification:** Make complex passages easier to read. * **Translations**: Translate the text into another language. * **Formatting:** Use the formatting toolbar to style your text (like bold, italic, underline, strike-through, or code format) or add links. ![AI\_Editor.gif](https://mintcdn.com/nuwacomgmbh/QmokFyROSwNjqeyr/images/AI_Editor.gif?s=afa200744b2a1cb33026174872edcbbd) **Ask Anything**\ Type your instructions in the input field to customize your text exactly how you need it. Examples: * **Adjust tone:** *'Make the text more engaging and lively.'* * **Shorten:** *'Shorten the text without losing its meaning.'* * **Add examples:***'Include three relevant examples in the text.'* ![AI\_Editor\_Ask\_Anything.gif](https://mintcdn.com/nuwacomgmbh/QmokFyROSwNjqeyr/images/AI_Editor_Ask_Anything.gif?s=e396c863737192a63bdae15c92998daa) \ \ Documents that you edit within projects are automatically saved. Nothing gets lost, even if you leave the page or pause your work. Once finalized, the text can be seamlessly copied and integrated into documents, presentations, websites, or any other relevant medium. ### **4. Organize documents in your team** Once you're working on multiple documents in a project, **statuses** and the **Kanban board** help you maintain overview and structure collaboration in your team. ![StatusandKanban.gif](https://mintcdn.com/nuwacomgmbh/EjzIIc2aC3Ah3CQt/images/StatusandKanban.gif?s=c29db361b90f5913896fccdff5dfcb94) You can assign a status to each document to make the editing progress transparent: * **Open** – Document has been created * **In Progress** – Editing in progress * **In Review** – Review by others * **Done** – Completed These statuses provide clarity about the current state of each document and facilitate collaboration. Projects can also be displayed as a **Kanban board**. There, all documents are sorted as cards by their status – so you can see at a glance what's open, what's being worked on, and what's been completed. **How to use the board:** * **Open board:** Click on *Board* in the top right of the Documents area to switch from table view * **Create documents:** Click on *+ New Document* in a column to create a document directly with the desired status * **Change status:** Drag and drop cards to another column or change the status directly in the document * **Open documents:** Click on a card to open the document. All changes are saved automatically * **Quick search:** Use search or filters to find documents by title, status, or assignee The Kanban board provides you with a visual overview of all ongoing work and supports agile workflows in your team. *** ## Develop ideas and conduct research in chat ![chatindocuments.png](https://mintcdn.com/nuwacomgmbh/-Mw3H6BZ6gOfZUuc/images/chatindocuments.png?fit=max\&auto=format\&n=-Mw3H6BZ6gOfZUuc\&q=85\&s=6e1f96e780a36aaa11862e3ead18898b) The chat on the left is your interactive workspace for collaborating with the AI.\ Here, you can brainstorm ideas, ask questions, or expand on existing texts — without leaving the editor. Use the chat to: * **Develop ideas and plan content** — e.g., topic suggestions for blog articles, social media posts, or campaigns. * **Submit research requests** — for example, on current trends, sources, or data from your **knowledge base.** * **Access agents** **specialized in specific tasks** (e.g., Fact Checker, Proofreader, or Social Media Writer). * **Refine or rephrase texts** while working in the editor simultaneously. The chat supports you throughout every stage of the writing process — from the first idea to the final polish. You can directly integrate results into your text or use them as a new foundation for further content creation. # Prompts Source: https://docs.nuwacom.ai/en/product/prompts Discover, create, and manage prompts to enhance your chat experience and simplify daily tasks. > [!NOTE] > > With the latest update, Prompts are disabled by default. Admins can re-enable them under Workspace Settings>Modules>Prompts. Once enabled, Prompts will appear again in the @ menu in chat and the navigation bar. ## **Prompt Library** The Prompt Library in nuwacom offers a central collection of proven prompts that can be executed with just one click — providing quick, targeted support for your daily work.\ You can choose between predefined prompts and custom prompts tailored to your specific use cases. **Standard Prompts** are predefined instructions developed by our team to automate common tasks efficiently and ensure consistency across the platform. Categories help you find the right prompt for various use cases: * **Brainstorming** – For ideation and creative processes. * **Email** – For standardized and professional email templates. * **External Comms** – For external communication such as press releases, blog articles, or website texts (like in the current view). * **Image Tools** – Image-related prompts, e.g., to generate image descriptions. * **Internal Comms** – Internal communication (e.g., team newsletters). * **Research** – For researching, structuring, and summarizing. * **Social Media** – For posts, captions, and campaign ideas. * **Text Tools** – Tools to optimize language, tone, style, etc. * **Archive** – For saved or older prompts. * *(Your own categories)* **Custom Prompts** allow you to precisely address individual requirements and specific use cases. Create your own templates, save them centrally in the prompt library, and make them available to your team. This article explains how to create, share, and use prompts step by step. *** ## **How to create prompts** ### **1. Go to the Prompts section** Open a new chat and enter “@”. Then hover over “Prompts” and click “Manage prompts”. This section is the starting point for creating and managing your prompts. Click the “Create prompt” button to create a prompt tailored to your individual workflow. > [!NOTE] > > By default, newly created prompts are private and only visible to you. Once you're done, you can share it with your team by clicking the share icon. ### **2. Enter title, category, and description** Provide a clear and descriptive title. Choose an appropriate category to keep your prompts organized and easy to find. Add a detailed description explaining the purpose and usage of the prompt. ### **3. Write the instruction** You have two options: * Write the instruction manually * Let the AI generate it **Write the instruction manually:**\ Define your expectations, assign a role to the AI, and describe the specific action. > [!NOTE] > > [**Prompt Engineering is a key part of working with generative AI models and helps you achieve precise and relevant results. Check out our article 🔗 “How to write effective prompts” for helpful tips.**](/resources/generateprompts) **Define variables for more precise results**\ Variables are especially helpful when creating your own prompts that require specific inputs. When a user selects a prompt, the defined variables appear in the input field and can be filled out individually, depending on the task. This ensures all relevant information is captured for accurate results. **Example – Variable:**\ Define variables by using curly brackets `{}` or by clicking the “Variable” button. ![promptvariableen.png](https://mintcdn.com/nuwacomgmbh/FtnmcdcRQjjMD_PP/images/promptvariableen.png?fit=max\&auto=format\&n=FtnmcdcRQjjMD_PP\&q=85\&s=7b2492b4fd9afd6e36b3b20bd22b20af) **Let AI generate the instruction:**\ Click the “Generate” button to let the AI create a detailed instruction. Simply provide a short description — ideal if you're short on time or unsure how to phrase your instruction. The AI will generate a structured and accurate prompt, improving the quality and efficiency of the interaction. ### **4. Select where the prompt appears** Choose the area in your workspace where the prompt should be prioritized: Editor, Chat, or Create. ### **5. Save the prompt** Once you've filled out all fields, click “Save.” > **Mark as Favorite** > > Use the star icon to add the prompt to your favorites and access it faster in the library. > **Share** > > Use the share icon to share the prompt with specific people or the entire workspace. > **Duplicate** > > The duplicated prompt opens directly in edit mode and can be adjusted — ideal for similar use cases. *** ## **How to create a Form** ![Promptform.png](https://mintcdn.com/nuwacomgmbh/S94SZUZSRDntcBTA/images/Promptform.png?fit=max\&auto=format\&n=S94SZUZSRDntcBTA\&q=85\&s=b18da546b7a46cc87e6d1d41b4821302) Display prompts as a form for a structured and repeatable execution of tasks.\ Forms are ideal when certain information needs to be collected in advance — such as the topic or target audience for a social media post. **How it works:** 1. Create a clear instruction for your prompt and add variables to capture the specific inputs you need. 2. Enable “Display as form” option to show the variables as fillable fields. 3. Prompt opens as a form: When users click the prompt, they’ll see a structured form. This ensures all required details are provided before the prompt runs. **Example – Form:**\ This form appears when the prompt “Internal Newsletter” is selected. The “Topic” input field is defined via a variable. ![formexample.png](https://mintcdn.com/nuwacomgmbh/FtnmcdcRQjjMD_PP/images/formexample.png?fit=max\&auto=format\&n=FtnmcdcRQjjMD_PP\&q=85\&s=027ded06c3219df5f9b296757c1c041f) *** ## **Where can I run prompts?** * **In the Prompt Library:** Execute prompts with a single click. * **In the Editor:** Highlight text, then use the "AI" button in the toolbar and select your prompt under “Prompts.” * **In the Chat:** Prompt suggestions will appear automatically as you begin typing in your chat. # Release Notes Source: https://docs.nuwacom.ai/en/product/releasenotes **New Update – October 23, 2026** # Design Systems for agents Design Systems can now be assigned to agents, so their output follows your brand’s colors, typography and design guidance without selecting a Design System in each chat. ## Permission before searching the web When web search is disabled and a request needs current information, nuwacom asks for permission before searching. Choose to search once, search and enable web search, or skip it. If you skip it, the AI answers using its existing knowledge, which may not be up to date. ## UI Optimizations ### A more compact navigation with Tools Apps, Agents, Skills, Design Systems, Workflows and Prompts are now grouped under **Tools** in the main navigation. This reduces the number of top-level entries and keeps the navigation easier to scan. Expand Tools to access all the same pages directly from the navigation. ## Connectors ### Improvements * **Google Sheets:** We’ve expanded the connector’s capabilities and improved its reliability. Find and update spreadsheet data, add or reorganise rows and columns, and manage tabs using spreadsheet and tab names instead of technical IDs. ### New integrations * **Google Analytics:** Connect Google Analytics 4 to let your agents analyze website traffic, identify top pages, compare periods and evaluate campaign performance directly in chat. Generate reports such as weekly traffic summaries, with read-only access to your analytics data. * **Canva:** Create editable designs from a brief, revise existing designs, add comments and export content with your agents. Available through Canva’s official MCP server, with admin setup requiring a custom Canva integration and credentials. * **Monday:** Find and update items, summarise board activity, and post comments directly from nuwacom. Filter work by owner, status, or due date to check progress and identify what needs attention. *** **New Update – October 16, 2026** # On-brand documents, slides and apps with Design Systems nuwacom can now create documents, presentations and apps using your brand’s colors, fonts, logos and design guidelines. The new Design Systems bring these details together, so you can reuse them without describing your brand in every prompt. **Build and customize your design system** Under **Design systems**, upload existing brand materials, such as a style guide or presentation, provide a website for the AI to crawl, or describe your brand in chat. The AI uses this input to build a design system you can review and edit. You can also build it manually by adding assets directly in the panel on the right. * **Colors and typography:** Define your brand palette, fonts and text styles in chat or add them directly. Font files can be uploaded separately. * **Logos and graphics:** Upload individual brand assets or extract them from your materials, where available in supported formats. * **Design guidelines:** Describe how these elements should be used and refine the guidance in chat or edit it directly. * **Multiple design systems:** Manage different brands or design directions and enable or disable each system individually. **Use your brand directly in chat** Select a design system through **@ > Design systems** and describe what you want to create. The AI can also use an enabled design system automatically. If several are available and you haven’t selected one, the new clarification dialog asks which to use. It can also clarify other missing details before creating the content, helping the result match both your brand and your requirements. **Fonts:** Freely available, license-free fonts, including suitable web fonts, can be retrieved automatically. Fonts that require a license must be uploaded separately. **Output quality:** Documents, slides and apps are AI-generated, with your design system guiding their appearance. Results may vary and may not match an existing layout exactly. We’re continuing to improve their quality and how precisely your design guidelines are applied. This release already brings a major improvement in quality, with documents, slides and apps that look great. ## Better answers and content with the new clarification dialog Missing context can make the difference between a generic result and one that fits your task. With the new clarification dialog, nuwacom asks for the information it needs before generating an answer or creating content. Your input helps improve the quality and relevance of the result. * **Clarify what’s missing:** When a request needs more context, the AI asks focused questions before proceeding. * **Give input quickly:** Choose a suggested answer or add your own details. * **Continue without answering:** Select **Skip** to proceed with the information already available. For on-brand documents and slides, this can mean confirming the design system or the required format. The dialog also works across other tasks, helping the AI understand what you need and deliver better answers and content. ## Add the right Knowledge faster We’ve redesigned the file selection in chat to make finding and adding context faster. Access recently used files, search for specific content or browse your Knowledge Base, all from **Files** below the chat input. * **Reuse recent files:** Recently used files appear when you open the menu, ready to add to the conversation. * **Search directly:** Find the files you need and add them from the search results. * **Browse your Knowledge:** Select **Browse all** to navigate your Knowledge Base’s folder structure and choose files or folders. ## Give Skills access to Web and Knowledge Search Skills can now search the web or your existing Knowledge when they need additional context. Enable Web Search or Knowledge Search in the skill’s settings. * **Use existing Knowledge without attaching files:** If the information is already in your Knowledge Base, enable Knowledge Search and describe in the prompt which content should be used as context. The AI can search for it without you attaching each file individually. * **Allow web research:** Enable Web Search so the skill can look up information online when needed. Knowledge Search uses the content you have access to in nuwacom. ## Switch models directly in the chat input The model selector has moved into the chat input, so you can select or switch models right where you describe your task. ## Connectors ### HubSpot We’ve simplified the HubSpot connector’s actions to make setup easier, improved search reliability and made error messages clearer. * Fewer actions to configure: Separate actions for individual record types have been consolidated. For example, Create Record now lets you select the record type, such as a company or deal, within one action. * More reliable searches: Find CRM information more reliably, with clearer error messages when something goes wrong. * Existing setups keep working: Your existing agents and workflows continue to work as before. To search logged emails, reconnect your HubSpot account once. ### Personio The new Personio connector brings employee information, absences and attendance into your work in nuwacom. * **Find employee information:** Look up employees and retrieve their details. * **Manage absences:** View, create and update absence records. * **Work with attendance:** Retrieve attendance information and record working time. ## New models * **Gemini 4 Argon:** Google’s latest reasoning model for complex coding, research and multi-step tasks. ## **Admin Controls** ### Make models available to the people who need them Admins can now control access to individual models under **Workspace Settings > Models**. A model can be made available to selected users or groups, for example power users who need it for specific tasks. * **Workspace access by default:** All workspace users are selected initially. * **Targeted access:** Click the settings icon for a model and select **Manage access** to restrict it to selected users and groups. ### Control access to documents, slides and emails by role The Permissions page now includes documents, slides and emails. Admins can configure which roles, including Admins, Editors and Readers, can read, create, edit and share each content type. *** **New Update – October 7, 2026** # **Start workflows directly from chat** Starting a workflow no longer means leaving the conversation. Instead of being redirected to the Workflows section, workflows now run directly in chat, with results appearing in the same place. To start, type @ and select a workflow from the menu. The workflow is attached to the message with *Run* already filled in. Press *Send* to run it, or replace *Run* with instructions to edit the workflow or a question about how it works. For workflows that use a form, the form appears directly in the conversation so it can be completed there. Once the workflow finishes, the result appears in the same chat, including links to any content it creates. You can then ask follow-up questions or continue working with the result. ## Improved connectors We’ve improved the reliability of actions in the Gmail and SharePoint connectors. Available actions include: **Gmail** * Find emails using subjects, senders, dates, previews, and attachment names. * Forward emails with their attachments. * Save attachments privately to your nuwacom Knowledge Base, individually or all at once. * Attach Knowledge Base files you have permission to access to replies, forwards, and draft replies. **SharePoint** * Read, create, and update pages, including adding sections or replacing content. * Save pages as drafts or publish them when ready. * Create folders and move or rename files and folders within the same document library. * Let your agent identify files and folders from SharePoint or OneDrive links you share in the conversation. * Check a file’s sensitivity labels to see how it is classified for handling. **New Update – October 1, 2026** # Quick Search: Find what you need, faster Finding the right chat, file, agent, or project should not slow you down. We’ve completely redesigned Search in nuwacom and turned it into a new Quick Search experience. You can now search across nuwacom from one place instead of navigating through different parts of your workspace. * **Search across nuwacom:** Find chats, files, agents, projects, apps, images, and more from one central search. * **Narrow down your search:** Switch between content categories or filter results by owner, date, project, type, or available actions. * **Preview content directly:** Hover over a search result to see more information and content without opening it. * **Return to recent work:** Access recently used items before starting a new search. * **Use multiple files in a chat:** Select multiple files and open them together in a new chat to use them as context. * **Find connected files:** Files from connected sources such as Google Drive that have been synced with nuwacom also appear in Quick Search and can be opened directly from there. * **Find common actions quickly:** Quick Search also gives you access to frequently used actions and navigation options. With Quick Search, finding and returning to the right content in nuwacom is now faster, easier, and much more intuitive. ## New models * **[GPT-6.1 Sol](https://developers.openai.com/api/docs/models/gpt-6.1-sol?utm_source=chatgpt.com):** OpenAI’s latest Sol model is designed for complex professional work while keeping costs lower than its flagship model, Astra. It is especially strong for tasks that require deeper reasoning, coding, or working through multiple steps and tools. * **Qwen3.8 Max:** Alibaba's most powerful Qwen model, designed for demanding tasks such as research, coding, and working with large amounts of information. It can understand text, images, and video. * **Qwen3.8 27B:** A versatile open model for everyday work, with particular strengths in coding and tasks that combine text and visual information. * **Qwen3.6 27B:** A strong open model for technical tasks, coding, and working with documents, images, and other visual content. * **MiniMax M3:** Designed for coding and more complex, multi-step tasks. It can work with very large amounts of information, making it useful for longer documents, larger codebases, and extensive context. All models listed above are available with EU hosting and need to be enabled by an admin in the Workspace Settings before they can be used. **New Update – September 30, 2026** # nuwacom is now on your phone! ![Release Mobile App Image](https://mintcdn.com/nuwacomgmbh/69GV6dmMff-oR4c_/images/release_mobile_app_image.webp?fit=max\&auto=format\&n=69GV6dmMff-oR4c_\&q=85\&s=f8f42e09945256fc62afa308422dc1a2) Your ideas don’t wait until you’re back at your desk. With the new nuwacom mobile app, you can capture them as they come, turn them into structured notes, and keep working with your AI workspace wherever you are. **Turn a thought into a structured note** Speak your idea into the Notetaker and create a structured note in just a few taps. Whether it’s a thought about a meeting, a question to explore later, or a possible solution to a problem you’re working on, you can capture it while it’s fresh. You can also open, edit, and add to notes you’ve already created. If you’re meeting others in person, you can use Start Recording to record the conversation with everyone’s consent. **Chat with the context you need** Start a chat and choose from different AI models. You can attach relevant Company Knowledge to bring information from your organization into the conversation, without having to gather and upload files each time. **Use your agents and skills on the go** Your agents and skills are also available in the mobile app. Use your agents for the tasks they were designed for, and use skills when you want to follow specific guidelines. **Keep your projects moving** Continue your work in your projects on the go: chat, create files, and start new projects for new ideas. You can keep working whenever you have time, wherever you are. With this first version of our mobile app, you can already use Company Knowledge in chats. Managing Company Knowledge and some other desktop features are not yet available on mobile. **Take nuwacom with you. Capture your thoughts and keep your work moving.** [Download for iPhone](https://apps.apple.com/de/app/nuwacom/id6766276903) · [Download for Android](https://play.google.com/store/apps/details?id=ai.nuwacom.app\&hl=de) ### Get started with the nuwacom mobile app [Download for iPhone](https://apps.apple.com/de/app/nuwacom/id6766276903) or [download for Android](https://play.google.com/store/apps/details?id=ai.nuwacom.app\&hl=de), then open the app and choose the option that matches your account. **Sign in to an existing nuwacom instance** 1. Select **Connect to private instance**. 2. Enter your instance domain. You can find it at the beginning of the URL you use to access nuwacom in your browser, for example, app-de.nuwacom.ai. 3. Tap **Save**. 4. Tap **Sign In** and enter your existing nuwacom credentials. **Create a new account** 1. Tap **Sign in** without connecting to a private instance. 2. Follow the steps in the app to create your account. 3. Sign in to start using nuwacom on your phone. *** ## Work in a folder from any chat A chat can now have its own working folder. Select a Knowledge Folder below the chat. If it already contains files, they are used as context for the conversation. Documents, presentations, and other files you create in the chat are saved in the same folder. This keeps your work in the chat together in one place. For example, you can use materials collected for a customer, discuss the next steps in the chat, and create a presentation based on what you’ve learned. The presentation is saved alongside the other customer materials, so you can return to it at any time or share it with your team. Select a folder below any chat to choose where you want to work. When you move to another topic, you can remove or change the selected folder at any time. *** ## Workflow performance at a glance The new Workflows overview shows how your automations are performing: total runs, success rate, and failed runs over the last 30 days. It also shows how many workflows you have created and published, with run information for each workflow in the list below. Publishing is easier to find, too. *Publish* now appears next to *Run* in the editor, and published workflows are marked Live in the overview. To take a workflow offline, select *Unpublish* from its three-dot menu. *** ## New models * **Claude Sonnet 5.5** is Anthropic’s fastest Sonnet model yet and uses fewer tokens per task than Sonnet 5. It is well suited to creating polished documents, slides, and spreadsheets, as well as coding tasks. [Learn more about Sonnet 5.5](https://www.anthropic.com/claude-sonnet-5-5). * **GLM 5.3 Flash** is an open-weight model with strengths in coding and understanding visual material such as charts and presentations. [Learn more about GLM 5.3 Flash](https://z.ai/blog/glm-5.3-flash). Both models are hosted in the EU. Please ask your workspace admin to enable them. *** ## For workspace admins **Configure permissions for the Admin role** Under Workspace Settings > Permissions, you can now enable or disable feature permissions for Admins, alongside Editors and Readers. This gives you control over which actions each role can perform across the workspace. *** **New Update – September 23, 2026** ## Slack: More reliable access and clearer interactions We’ve improved how you can access, search, and work with Slack content in nuwacom. * **Access the Slack content available to you:** The Slack connector now uses your individual Slack login, giving you access to the same channels and messages through nuwacom that you can access in Slack. * **Find the right channels reliably:** Channel search has been fixed, making it easier to locate channels by name. * **Understand messages more easily:** Slack results now display real names and dates instead of technical identifiers. Results are also more concise and faster to process. * **Stay in control when posting:** Before sending a message, nuwacom asks if it should be posted. Replies always require your confirmation. These improvements make it easier to find relevant conversations, summarize discussions, follow message threads, and prepare or send Slack messages. ## **Pipedrive**: Connection issues resolved We’ve resolved the previous connection issues with the Pipedrive integration, making it available again. With Pipedrive connected, leads, deals, contacts, notes, and activities can be created, updated, and retrieved. This supports workflows such as: * Looking up customer and deal information * Creating new leads and contacts * Updating deals and activities after meetings or follow-ups * Adding notes to the relevant CRM records * Reducing manual data entry across recurring sales workflows ## UI updates We’ve made several interface improvements to create a clearer and more consistent experience across nuwacom: * **A clearer @ menu:** We’ve unified the menu, removed unnecessary labels, and updated its icons and avatars, making the available options easier to scan and understand. * **Consistent avatars:** Apps, agents, and skills now use aligned and updated avatars, making them easier to recognize across nuwacom. * **Unified chat selections:** Agent, voice, and skill chips now follow the same visual pattern, making selected options easier to identify. * **A more structured model overview:** Models are now grouped by provider on the Models settings page, replacing the previous long list with a clearer and easier-to-navigate overview. *** **New Update – September 16, 2026** We've made notes, knowledge, and data protection work harder for you. Here's what's changed. ## Notetaker: A new home for your notes and meetings The redesigned Notetaker brings notes and upcoming meetings together in one clear overview, making it easier to prepare for calls and revisit previous conversations. * **See what's next at a glance:** Upcoming meetings are visible at a glance, with direct control over which ones the Notetaker bot joins. * **Join any meeting:** Paste a meeting URL to send the bot to calls outside the connected calendar, including meetings hosted on Zoom or Webex. * **Capture a note in one click:** Start a blank note and type, or hit record — the conversation is transcribed and summarized automatically. * **Find notes faster:** Search notes or filter them by people, sources, creation date, and modification date. * **Start a chat from notes:** Select one or multiple notes and use their content as context for a new chat. ## Knowledge: All your knowledge, easier to find and use The Knowledge Base now gives a clearer picture of everything connected and worked on — and more ways to act on it without leaving the page. * **All sources, visible at a glance:** Data Connectors (like SharePoint) and synched spaces and folders (from Confluence or Google Drive) appear directly in the sidebar. * **Create without switching pages:** Start a new document, slide, note, email, image, or video directly from the Knowledge. * **Do more with less clicking:** Select multiple files to start a chat, move, share, or download them together — or start a chat straight from a single file. * **Find anything instantly:** Filter by type, source, date, person, or project, and search across the entire Knowledge Base. ## Privacy Mode: Keep Sensitive Data Protected by Default ![Privacymodenenglish](https://mintcdn.com/nuwacomgmbh/gIBh18us90pfbdbm/images/privacymodenenglish.png?fit=max\&auto=format\&n=gIBh18us90pfbdbm\&q=85\&s=c511d4a0802c883fc74847e1f5ca3cb3) The new Privacy Mode automatically protects personal data before it ever reaches the AI. * **Anonymize data before processing:** Recognized information — such as names, contact details, customer and employee data, financial information, contract details, and other confidential business data — is replaced with neutral placeholders before it reaches the AI model. * **Protect prompts and files:** Anonymization applies to both chat input and uploaded documents. * **Stay informed:** Anonymized files are clearly labeled, and protected values appear as placeholders in conversations. * **View original information when needed:** Hover over a placeholder or switch between the anonymized and original display within nuwacom. Enterprise customers can try Privacy Mode at no additional cost until the end of the year. To enable it for your organization, contact your Customer Success Manager. ## Build with nuwacom: Agents & Skills APIs The new public Agents and Skills APIs let you create and manage agents and skills from your own tools, scripts, and systems. * **Create and update agents and skills:** Configure an agent’s name, instructions, model settings, and actions in a single request. Create and edit skills and manage their supporting files. * **Work safely with drafts.** Changes never go live by accident: you edit a draft and *publish* it when it's ready, just like in the app. Unpublish any time to take an agent offline without deleting it. * **Switch skills on and off** for a whole space with one call, no need to delete anything. Seasonal playbook? Enable it in October, disable it in January. * **Chat with agents:** Use the OpenAI-compatible Completion API to chat with any agent. Its full configuration, including instructions, knowledge, and model, is applied automatically. * **Stay in control.** Everything runs through workspace API keys with fine-grained permissions, scoped to a space. What an agent can see and who can use it stays governed by your workspace settings. The Agents & Skills APIs are now available in all workspaces. Admins can create an API key in the workspace settings. Explore the [Agents API documentation](https://docs.nuwacom.ai/admin/api/agents-overview) and [Skills API documentation.](https://docs.nuwacom.ai/admin/api/skills-overview) ## Additional Connectors * **Microsoft OneNote integration:** Connect OneNote to bring existing notebooks and notes into the AI-powered work context. * **Microsoft Planner integration:** Connect Planner to make plans and tasks accessible alongside the rest of the work context — so planning sits next to other work instead of living in a separate tool. ## New Models * **GPT-6 Astra:** OpenAI’s new model for complex tasks, designed to reliably maintain context across many steps. * **Gemini 3.8 Flash:** Google’s most intelligent Flash model delivers fast, reliable results even for complex workflows. * **DeepSeek V4.1 Flash:** DeepSeek’s latest model analyzes both text and images and delivers fast, reliable results, even for complex tasks. * **GPT Images 2.5:** This new image model delivers sharper details, more precise edits, faster generation, and better consistency across multiple revisions. All models listed above are available with EU hosting. ## Small improvements, big impact * **Permission updates:** Admins can now control whether users can create user groups and whether chats, documents, skills, agents, and other content can be shared with the entire workspace. * **Increased Skill limit:** The maximum number of Skills has been increased to 40. * **Agent Analytics:** Analytics now supports filtering by active agents. *** **New Update – August 27, 2026** # Sovereign open-weight models: strong performance for European AI independence nuwacom now includes Kimi K3, DeepSeek V4, and GLM 5.2. These models expand the available model portfolio with leading open-weight options that support a more sovereign AI strategy in Europe: strong performance, lower cost, and less strategic dependence on US providers. ### What’s new * **Leading open weight models are now available in nuwacom**\ Kimi K3, DeepSeek V4, and GLM 5.2 can now be selected directly in nuwacom workspaces. All models are exclusively running inside the EU on infrastructure operated by European companies. * **A more sovereign European AI setup**\ The new models support strong AI performance with less dependence on US providers. In combination with nuwacom’s STACKIT partnership, they support a setup in which sensitive data stays within the EU. * **Strong quality with better cost efficiency**\ Open weight models now offer leading performance at significantly lower cost than many proprietary alternatives. ![Nuwacom Llm Stats Top10 Scores Costs En](https://mintcdn.com/nuwacomgmbh/n4lKCOY9l7_aCUKc/images/nuwacom-llm-stats-top10-scores-costs-en.png?fit=max\&auto=format\&n=n4lKCOY9l7_aCUKc\&q=85\&s=59fd383d9ec2fa04e63b6536d42467d8) ### Why this matters For a long time, organizations had to choose between top model performance and a sovereign European setup. That trade-off has become much smaller. With the latest generation of open-weight models, strong quality, lower operating cost, and European sovereignty can now come together in one setup. This gives teams more flexibility in how they build their AI stack and reduces long-term dependence on US hyperscalers and AI providers. [Learn more.](https://nuwacom.com/en/resources/blog/leading-open-weight-models-cost-effective-path-sovereign-ai-europe) ## **Agents and Skills: templates and a more intuitive way to create** Agents and Skills now include templates that offer a faster starting point and make creation more intuitive. Their overview pages have also been reworked and aligned to create a clearer and more consistent experience across browsing and navigation. ### **What’s new** * A selection of templates for different departments provides inspiration and a faster setup. * The overview pages now follow a more consistent structure across layout and navigation. ## Small improvements, big impact * Admins can now control which roles have which permissions for the Notetaker module. * The Recent list in the navigation and the “Show more” view have been made consistent and now feel aligned. * The Apps overview page now provides a clearer browsing experience. *** > **Archive** > > **New Update – July 31, 2026** > > # Workflows can now create documents, emails, and slides > > Workflows can now turn results into documents, email drafts, and slide decks. You can save them directly to the knowledge or a project. > > #### What’s new > > * **Create content:** Create documents, email drafts, and slide decks in a workflow. > * **Choose how content is created:** Use AI to create new content or use the result from an earlier step. Skills help keep the right tone and format. > * **Create and save several results:** Create multiple items in one workflow and save them to a folder in the knowledge or a project. > > #### Why this matters > > * **Get content you can use right away:** Teams can review, edit, and share the created content. > * **Save time:** Create several items in a single workflow run. > * **Keep everything organized:** Save each item in the right knowledge folder or project. > > ![Releasenotesworkflows 1](https://mintcdn.com/nuwacomgmbh/iv1Cu73TG30JH6P1/images/releasenotesworkflows-1.png?fit=max\&auto=format\&n=iv1Cu73TG30JH6P1\&q=85\&s=1387a3ff5d4870982489fcbce6f08d94) > > #### How it works > > 1. Add an output step to a workflow. > 2. Select the type of output to create. > 3. Choose whether the step should generate new content with AI or use an earlier workflow result. > 4. Define where the output should be saved. > 5. Add further output steps if more than one deliverable is needed. > 6. Run the workflow. > > ## Small improvements, big impact > > * **Optimized Chat Experience** \ > The chat experience has been optimized and is now easier to follow, more clearly structured, and overall more user-friendly. > * **Improved Source Display** \ > Sources are now shown directly, making the underlying references easier to understand.In-App Notification > > *** > > **New Update – July 17th, 2026** > > ## Build apps with your own brand and visual assets > > ![App Style 1](https://mintcdn.com/nuwacomgmbh/dVaJ7QSQcGjawpQi/images/App_Style-1.png?fit=max\&auto=format\&n=dVaJ7QSQcGjawpQi\&q=85\&s=8f9e96d84223239656f83155c6e93109) > > Apps can now be created with brand and style inputs directly in the App Builder. Visual assets and styling information can be included from the start, so the generated app reflects the intended look and feel more consistently. > > This update makes it easier to build production-ready apps that combine functional structure with the right visual direction. > > ### What’s new > > * **Use brand and style inputs during app creation**\ > App creation now supports styling information and visual assets as part of the build process. > * **Add uploaded images to the app build**\ > Uploaded images can now be used directly when creating an app. > * **Apply design guidance through a skill**\ > A design skill can be attached during app creation to transfer styling information such as typography, colors, logos, images, and other relevant assets. > * **Refine the styling after the app is built**\ > Once the app is generated, the styling can still be adjusted through the chat by providing additional design instructions. > > ### Why this matters > > This update helps close the gap between an initial idea and a shareable, production-ready app experience. > > It supports use cases such as: > > * building landing pages with visuals and consistent brand styling > * creating campaign pages from a design brief with product screenshots and approved images > * developing sales tools with proposal configurators and company-branded microsites > * turning a presentation, concept document, or image set into an interactive app or dashboard > * creating internal tools with the right logos, illustrations, and visual style from the start > * building event pages, product launch pages, or executive demos with a stronger design standard > > ### How to use it > > 1. Select the **Build App** use case in the chat. > 2. Describe the app you want to create. > 3. Add styling information directly in the chat, and upload any assets you want to use. > 4. Alternatively, attach a design skill that contains the relevant design guidance, such as typography, colors, logos, images, and other assets. > 5. Generate the app. > 6. Refine the styling afterward in the chat if further adjustments are needed. > > *** > > **New Update – July 3rd, 2026** > > #### New Feature > > ## Apps: build your own tools directly in Chat > > ![Image](https://mintcdn.com/nuwacomgmbh/xhfIQQsia-TNL1Zs/images/image-4.png?fit=max\&auto=format\&n=xhfIQQsia-TNL1Zs\&q=85\&s=27aa7ed259153e8ae6e99796628a904b) > > We've all been there: you chat with an AI, get a great text response, and then... you copy-paste it into Excel, manually build a chart, or format it into a table. > > With Apps, that changes. You're no longer just generating static text - you're building real, interactive, reusable tools directly inside your chat, without writing a single line of code. Need a custom pricing calculator for your sales team, a live project dashboard, or a structured feedback form? Instead of waiting on IT or wrestling with complex software, you build it yourself in a few minutes. > > **What you can build** > > * **Dashboards & reports:** KPI overviews, revenue trackers, signup/conversion funnels, or project status boards. > * **Tools & calculators:** Custom pricing configurators, ROI calculators, or internal helper tools. > * **Forms & intake panels:** Support intake forms with file uploads, settings panels, or custom filters. > * **Knowledge explorers:** Searchable, interactive interfaces built on top of your company's knowledge base. > * **Boards & maps:** Visual strategy boards, process flows, or customer journey maps. > > **How it works, in three steps** > > 1. **Launch the builder.** Type @ in your chat input field, select Create from the menu, and click “Apps”. > > 2. **Describe your app.** Select a suggested prompt or describe in Chat what you want to build, how it should work, and how it should look (e.g. "Create a project tracker with a status dropdown, a priority filter, and a clean, dark-blue layout"). nuwacom builds the actual interface for you - buttons, input fields, charts and all. > > 3. **Interact & share.** Test your app immediately in the preview. Once it works exactly how you want, share it with your team. > > > [!NOTE] > > > > **Want to try it?** The Apps feature is rolling out gradually. If you don't see the Apps tab in your sidebar yet, reach out to your internal workspace administrator to have it enabled for your organization. > > **Want to learn more?** Read our official [announcement post](https://nuwacom.com/en/resources/blog/apps-build-your-own-business-software-inside-nuwacom) and find out more about the vision behind Apps in [this post](https://nuwacom.com/en/resources/blog/the-end-of-software-you-have-to-obey) by our founder & CEO Christophe Folschette. > > #### New Feature > > ## Workflow Triggers: let nuwacom start tasks for you automatically > > With this update, nuwacom can start a workflow for you automatically, the moment something happens in your workspace. For example, a workflow can run on its own as soon as a new meeting note is created, or as soon as a knowledge base file has finished processing. > > You set this up once, and the workflow takes over from there. > > * **Start workflows automatically when something happens.** Choose an event inside nuwacom - such as a new meeting note being created - and the workflow runs on its own when that event occurs. > * **Run them only when it makes sense.** Add filters so a workflow runs only under specific conditions - for example, only when a file is added to a particular Knowledge Base folder, rather than every folder. > * **Carry information from the event into the workflow.** Details from whatever triggered the workflow (the meeting note, the file) are passed straight into the following steps, so the workflow and any AI agents already have the context they need. > * **Set it all up in one place.** When you add or edit a trigger in the workflow builder, you'll see a new option to choose a nuwacom event, such as "Meeting Note Created" or "Knowledge Base File Processed." If filters are available for that event, they'll appear right there in the same panel. > > No external connections, webhooks, or third-party tools are needed - everything runs inside nuwacom. > > *** > > #### New Model > > ## Claude Sonnet 5: pick it for your most demanding work > > ![Image](https://mintcdn.com/nuwacomgmbh/xhfIQQsia-TNL1Zs/images/image-5.png?fit=max\&auto=format\&n=xhfIQQsia-TNL1Zs\&q=85\&s=bd1bc08860eea6a1c9e5d56fc4bdeaf3) > > Claude Sonnet 5 is now available in nuwacom as an EU-hosted model. Choose it for any chat when you need to think through a hard problem, work across several steps, or use tools > > * **Reach for it when work gets complex** - multi-step tasks, harder reasoning, tool use. > * **Dial the effort to fit the task** - turn it up for depth, down for speed. > * **Trust the output more** - fewer incorrect or made-up answers than before. > > *** > > #### Update > > ## Optimized Navigation: we've streamlined your sidebar to fit the way you work > > ![Group 14735](https://mintcdn.com/nuwacomgmbh/OCtxFWSkjhEcCSyM/images/Group-14735.png?fit=max\&auto=format\&n=OCtxFWSkjhEcCSyM\&q=85\&s=2d2daaf608bc9c849ceee123402a379a) > > \ > We've reorganized your sidebar to put the things you use closer to where you work. Here's what's new. > > * **"Meeting Notes" is now "Notetaker."** It has been renamed and remains directly accessible in your sidebar navigation. > * **Create Slides, Documents, Apps, Images, Videos and more are now available in Chat.** They are no longer in the sidebar - type @ and select Create to find them. > * **Manage prompts now directly in chat.** They are no longer in the sidebar - type @, hover over Prompts, and click Manage Prompts or choose the one you want to use. > * **Apps is a new section in your sidebar.** It's a new entry in its own right (see the Apps section above) - not a renamed or moved version of anything that was there before. If you don't see the Apps tab in your sidebar yet, reach out to your internal workspace administrator to have it enabled for your organization. > * **Documents in the Knowledge Base:** From now on, all your files are centrally available in the Knowledge Base. Use the file type filters to quickly find the file format you're looking for. > > *** > > **New Update – June 26, 2026** > > #### File Previewer > > ## **See what nuwacom creates - instantly** > > Just click the generated file to open an in-app preview. From there you can: > > * View the file instantly in a preview > * Open it in full screen > * Download it with one click > > **Why this matters** > > * Stay in your flow - no app switching > * Quickly verify the output before you use or share it > > *** > > #### Skills with assets > > ## **Add your own files to Skills** > > Skills define how nuwacom responds - the tone, structure, and format of what it generates. You can now attach **assets** directly to a Skill, so the Skill can work with *your* materials (for example, brand files and references). > > **Supported formats you can add to Skills** > > * **Text:** md, txt, csv, html > * **Images:** svg, png, jpeg/jpg, webp, gif > * **Fonts:** woff, woff2, ttf, otf > > **Why this matters** > > * Keep outputs consistent by grounding a Skill in your own files and references > * Reuse common building blocks without re-uploading or re-explaining them every time > > *** > > #### Gemini Flash 3.5 > > ## **Faster instant model in Auto mode** > > In Auto mode, nuwacom automatically selects the best model for your request. The **instant model** (used for quick, direct responses) has been updated from **GPT 5.2** to **Gemini Flash 3.5**. > > **What you’ll notice** > > * Faster response times, so quick questions feel even more instant > * Stronger overall quality based on improved benchmarking > * More room for context, so longer inputs are handled more reliably > > *** > > ### **Small improvements, big impact** > > **Meeting transcript export:** Copy or download the transcript of any meeting note directly from the transcript view. > > **Meeting Bot name in meetings:** The Meeting Bot now automatically uses your first name, so it shows up with a recognizable name instead of a generic one. > > > [!NOTE] > > > > **Note:** If you already have an Meeting Agent set up, disconnect it once and then add it again. This is required for the Meeting Bot to appear in meetings using your first name. > > *** > > **New Update – June 17, 2026** > > # **Memory - nuwacom now remembers more reliably what matters to you** > > ![Group 14713](https://mintcdn.com/nuwacomgmbh/yxhlAkg-05w-SRtG/images/Group-14713-1.png?fit=max\&auto=format\&n=yxhlAkg-05w-SRtG\&q=85\&s=4957b39b6cc2bf5bc3281dfe8c6c4739) > > Your Personal Memory has been redesigned. Information is now reliably filtered and stored - nuwacom keeps track of who you are, how you work, and what you need. Precise answers without describing the same context over and over. > > ### What's new > > * Full context on every turn: nuwacom pulls information from your memory whenever it's relevant. > * Organised into clear sections: Your memory is now split into categories like Profile, Preferences, Goals, and Decisions - so you can see at a glance what nuwacom knows about you. > * Update your memory anytime: Tell nuwacom what to remember, change, or forget directly in chat - or manage entries manually in your settings. > * Corrections that actually work: "I no longer live in Berlin" now replaces the old entry instead of creating a contradictory one alongside it. > * Import memory from other AI tools: Bring over existing memory from tools like ChatGPT or Claude and continue using it in nuwacom. > > [→ Learn more](/product/accountsettings#memories) > > ## **Auto Mode - nuwacom chooses the right model for you** > > ![Group 14710](https://mintcdn.com/nuwacomgmbh/yxhlAkg-05w-SRtG/images/Group-14710-2.png?fit=max\&auto=format\&n=yxhlAkg-05w-SRtG\&q=85\&s=1700e2c8386cd870f9d7317d26e55c26) > > With Auto Mode, you no longer need to worry about which model is the best choice for each task. Auto Mode automatically selects the most suitable model for every task, so you can focus on the work instead of comparing models. > > ### What's new > > * Automatic model selection: nuwacom chooses the right model for each message, EU-hosted by default. > * Faster starts: No need to decide upfront which model to use. > * More efficient usage: Simple requests use faster models; complex tasks get routed to the bigger models. > * Always up to date: We regularly update the models behind Auto Mode, so you benefit from the latest improvements without keeping track yourself. > > From today on, new chats start in Auto Mode by default. You can set a personal default model in your preferences - but Auto Mode is usually the better choice because it's more efficient, ensures high output quality, and is always up to date. > > [→ Learn more](/product/chat#model-selection) > > ## **Meeting Notes - add Zoom and Webex meetings by link** > > ![Group 14726](https://mintcdn.com/nuwacomgmbh/yxhlAkg-05w-SRtG/images/Group-14726.png?fit=max\&auto=format\&n=yxhlAkg-05w-SRtG\&q=85\&s=95e9858af03519794a926f0ad257e817) > > In addition to connecting your Outlook or Google calendar, you can now add Zoom or Webex meetings by pasting the meeting link directly. The bot records, transcribes, and summarizes everything - so you can focus fully on the conversation and use the notes afterwards for follow-ups, debriefings, or as context for your next task. > > [→ Learn more](/notes#add-notetaker-via-meeting-link) > > ## **TL;DR** > > **Personal Memory** now remembers more reliably what matters to you - with a clearer format, chat-based updates, better corrections, fewer duplicates, and import from other AI tools. > > **Auto Mode** is now the default for new chats and automatically chooses the right model for each message - always fast, efficient, and up to date. > > **Meeting Notes:** Zoom and Webex meetings can now be added by pasting the meeting link directly into nuwacom. > > *** > > **New Update – June 4, 2026** > > ## **Granular Permissions - full control over what each role can do** > > Admins get a powerful new way to manage access through the new *Permissions* settings. Instead of simply deciding *whether* a role can access a feature, you can now define exactly *what* people can do inside it - view, edit, and share. > > * **Control per role *and* per feature:** For each feature, define granular rights for Editors and Readers - View, Edit, and Share (where applicable) - instead of a simple on/off switch. > * **Covers your whole workspace:** Agents, Orchestrator, Workflows, Projects, Skills, Prompts, Voices, Image and Video Generation. > * **Enforced instantly, everywhere:** Changes apply in real time. If a role doesn't have access, the feature simply isn't available to those users - it's hidden from their menus and can't be reached through a direct link or any workaround. No logging out and back in. > * **Works alongside sharing:** Permissions decide which features a role can use at all, while your familiar sharing settings still control access to individual content. The two work hand in hand. > > **Why it matters:** You get precise, reliable controls over your workspace - exactly the right access for every role, applied automatically and without gaps. > > *** > > ## **The Superagent is now the Orchestrator - your AI that coordinates work for you** > > Say hello to the *Orchestrator* - a powerful new way to get things done without picking the right agent or skill yourself. Just describe your task, and the Orchestrator brings together everything needed to complete it. > > * **Easy to find:** The Orchestrator now appears as a button right under the chat input and is listed under AI Apps in the left navigation. > * **Suggested prompts to get you started:** When you select it, you'll see ready-made example prompts that show what the Orchestrator can do and how to put it to work - so you're productive from the first click. > * **It coordinates everything for you:** The Orchestrator automatically pulls in the agents and skills your task needs, step by step - without you naming them. Ask for research, and it brings in the Research Agent on its own. > > ![Orchestratoren](https://mintcdn.com/nuwacomgmbh/EZbccf0IzaHaONyB/images/orchestratoren-1.png?fit=max\&auto=format\&n=EZbccf0IzaHaONyB\&q=85\&s=5b0e605332ab62d66fb2c380bce23279) > > *** > > ## **Find and trigger everything faster in the chat - clearer + and @ menus** > > * **Click + to add content and choose where to search.** Use it to bring in context (files, knowledge, content from Apps) and to pick your search sources (Web, Knowledge Base, and connected sources like Teams or Confluence). > * **Type @ to reach all your AI in one place.** Just type **@** while writing, and a menu opens with your AI Apps, Agents, Skills, Prompts, Voices, and Workflows. > > **Why it matters:** You always know where to go - **+** to add and search, **@** to call on your AI - so getting started is faster. > > ![Aimenuen](https://mintcdn.com/nuwacomgmbh/FBUzIFKQ_xmikLDJ/images/aimenuen-1.png?fit=max\&auto=format\&n=FBUzIFKQ_xmikLDJ\&q=85\&s=db7361f5e0834c371768d81e66e58aa6) > > *** > > ## **TL;DR** > > * **Granular Permissions:** Admins can now define exactly what each role can do per feature - view, edit, and share - not just whether they have access. Applies across your whole workspace and takes effect instantly. > * **Orchestrator:** A new way to get things done that coordinates the work for you. Just describe your task, and it brings in the right agents and skills on its own - no need to pick them yourself. > * **Clearer + and @ menus in the chat:** **+** to add content and choose where to search, **@** to reach all your AI in one place. You always know where to go. > > *** > > **New Update – May 20, 2026** > > # New Image and Language Models > > We expanded the nuwacom model catalog with 3 new image models and 6 language models. > > This update also strengthens our **open-source model offering** with the addition of gpt-oss-120b. > > This gives you more choice when selecting the right model for your task — whether you need high-quality image generation, image editing, fast answers, more complex analysis, open-source workflows, multimodal work, or long context handling. > > ## New image models for image generation > > With this update, FLUX models from Black Forest Labs are available - depending on the model, for high-quality image generation, editing existing images, or creating more complex image ideas. > > * **FLUX 1.1 \[pro] Ultra** For fast, high-resolution, and detailed images. Especially useful when you want to create a high-quality single image from a prompt with flexible aspect ratios. > * **FLUX1 Kontext** For editing and further developing existing images. You can use an image as a reference and adjust it with a text prompt, for example to change the style, background, or individual image elements. > * **FLUX2 \[pro]** For more complex image ideas with many details. The model follows more detailed descriptions more accurately and is especially useful when the subject, style, composition, or text in the image should match your prompt as precisely as possible. > > 👉 Image models can now also be enabled or disabled by admins in the settings. > > #### Image model selection in Create Image > > In the *Create Image* task, you can now choose the image model before generating an image. > > This helps you use the model that best fits your image task — for example, if you want to create especially detailed images with the new FLUX 1.1 Ultra model. > > ![Image Modelselection](https://mintcdn.com/nuwacomgmbh/zTJWrbWaazxUfYuO/images/ImageModelselection.png?fit=max\&auto=format\&n=zTJWrbWaazxUfYuO\&q=85\&s=8777b1c61072b7649bd05809af9ace94) > > ## New language models > > The language model catalog has also been expanded with new models for reasoning, fast responses, open-source workflows, multimodal tasks, and long context handling. > > Depending on the hosting region, the following models are now available. > > #### Europe-hosted models > > * **GPT-5 Thinking / GPT-5 Thinking Fast** For more demanding tasks where the model should reason through relationships more carefully — for example in analysis, planning, coding, or more complex questions. Thinking Fast is the faster option if you want a good balance between quality and response time. > * **GPT-5.1 Thinking / GPT-5.1 Thinking Fast** The updated GPT-5 generation for complex tasks, coding, and multi-step workflows. Suitable when you need more precise results for tasks that involve several steps or decisions. Thinking Fast is useful for everyday tasks where you still benefit from stronger reasoning but want a faster response. > * **o3 Mini High** For technical and logical tasks, especially in coding, mathematics, analysis, and problem-solving. A good choice if you are looking for a compact model with strong reasoning capabilities. > * **Gemini 2.5 Flash** For fast responses and efficient everyday tasks. Especially useful for summaries, short analyses, quick text work, and tasks where speed matters more than maximum depth. > * **Gemini 2.5 Pro Reasoning** For more complex analyses, coding, and tasks with a lot of context. Suitable when the model needs to understand and process broader relationships more carefully — for example with longer documents, technical questions, or multi-step tasks. > * **gpt-oss-120b – Open Source** A large open-weight model from OpenAI for more demanding text, coding, and analysis tasks. Suitable if you want a powerful open model with strong problem-solving capabilities. > > ### TL;DR > > nuwacom is expanding the model catalog with 3 new image models and 6 language models for image generation, image editing, fast responses, more complex analysis, open-source workflows, multimodal tasks, and long context handling. > > In the *Create Image* task, you can now also choose the image model before generating an image, making it easier to select the model that best fits your task. > > *** > > **New Update – Apr 29, 2026** > > ## **Skills – Define how nuwacom response, once** > > ![Skillsen](https://mintcdn.com/nuwacomgmbh/Z4t0Ls5BZHtPHQhF/images/skillsen.png?fit=max\&auto=format\&n=Z4t0Ls5BZHtPHQhF\&q=85\&s=e5684a5e4344883179fb729009afb86c) > > A Skill defines how the AI responds in chat — including tone, structure, rules, and output format. > > You create a Skill once instead of repeating the same instructions every time. It then runs automatically in the background: when you start a chat, nuwacom reads the descriptions of all active Skills, determines which ones are relevant, and applies them directly — without you needing to select or mention anything. > > This makes a Skill a reusable standard that ensures nuwacom works consistently and reliably every time. > > ### **What’s new** > > * **Create reusable Skills** to standardize how nuwacom responds across recurring use cases like competitor analyses, briefs, reports, customer responses, anything you produce in a defined format. > * **Automatic Skill activation** in the background. nuwacom recognizes the prompt and applies the right Skill without you needing to remember which one to use. > * **Combine Skills** to layer tone, structure, and formatting rules in a single output. For example, a "Brand Voice" Skill plus a "Technical Brief" Skill working together. > * **Workspace sharing** so your whole team uses the same Skills. > * **Export and manage** Skills as structured files (.md, .zip, .skill) > > ### **Skills in Agents** > > Skills can be attached directly to an Agent. When attached, the Skill is active for anyone using that Agent, without requiring each user to install it. > > Picture a Competitor Analysis Agent with the matching Skill built in. A team member opens it and asks: > > *“Analyze how our competitors position themselves in the market.”* > > The output comes back already structured the way your strategy team defined it including the same sections, same depth, same standards. Whether the request comes from marketing, sales, or product, the analysis follows one consistent format. > > This means you can ship fully configured Agents to your teams with tone, structure, and format already locked in. > > ### **Skills vs. Agents — The Critical Distinction** > > | | **Skill** | **Agent** | > | :- | :- | :- | > | **Defines** | How nuwacom responds | What nuwacom does | > | **Focus** | Format, tone, structure | End-to-end task execution | > | **Mode** | Passive — activates automatically or on demand | Active — can run autonomously | > | **Think of it as** | The rulebook | The executor | > > ### **New to Skills?** > > Use the Skill Creator to build your first Skill through a simple conversation with nuwacom without any complex setup. Describe what you need, and nuwacom turns it into a reusable Skill in minutes. > > 👉 [Learn more](/product/skills) > > ## **Model Upgrades: Frontier Intelligence and sharper Videos** > > **Three new models join nuwacom: GPT-5.5, Claude Opus 4.7, and GPT-Image 2 alongside upgrades to video and Gemini.** > > We're keeping nuwacom's model layer at the frontier, so your teams always work with the strongest, most reliable AI models available without ever leaving your secure EU environment. > > * **GPT-5.5 by OpenAI**: OpenAI's newest frontier model, delivering meaningful gains in agentic coding, computer use, and complex knowledge work for the most demanding enterprise tasks. > * **Claude Opus 4.7 by Anthropic**: Anthropic's most advanced model, built for nuanced reasoning, long-context analysis, and high-quality writing where output quality matters more than anything else. > * **GPT-Image 2 by OpenAI**: OpenAI's state-of-the-art image model with markedly better text rendering, multilingual layouts, and instruction following. Production-ready for slides, infographics, and on-brand visuals. > * **Veo 2 (EU)** is now the default for video generation, replacing the deprecated Sora 2 with sharper output and more natural motion. > * **Gemini 3.1 Pro** replaces Gemini 3 Pro, bringing stronger reasoning and improved tool use. > > ## **Small improvements, big impact** > > **Background streaming in chat:** Chat responses continue streaming even when you navigate to another part of the app, so you can keep working without waiting. > > **Cleaner chat messages:** Improved message formatting makes responses easier to read and scan. > > **Expanded template dialog:** Preview templates more easily with an improved dialog experience. > > **Consistent chat experience:** New chats start with your last-used knowledge sources and AI model. > > **Updated navigation:** “Meeting notes” is now renamed and pinned to the sidebar by default for quicker access. > > *** > > **New Update – Mar 24, 2026** > > ## **From chat to getting work done** > > ![01133685 5359 4d90 81f8 C563f60f07cd](https://mintcdn.com/nuwacomgmbh/msQ67eH024EqkZjX/images/01133685-5359-4d90-81f8-c563f60f07cd-2.png?fit=max\&auto=format\&n=msQ67eH024EqkZjX\&q=85\&s=e0813e1345f6ba3038a555b63f9cd1ef) > > We’ve been thinking a lot about how people actually use AI at work. > > Most tools still start with a chat. You ask something, you get an answer — and then the real work begins somewhere else. > > That’s exactly what we wanted to change. > > This release marks a major step forward moving from nuwacom 1.0 to nuwacom 2.0, and from AI that answers to AI that actually gets work done. > > With nuwacom 2.0, we built a centralized AI workspace that knows your company and gets work done. > > Now you can create, refine, and manage real work- from documents and slides to workflows and notes - all in one place. > > ## **From chat to tasks to real work** > > In nuwacom 2.0, you still have your chat and you can choose the model you want to work with. But now, it goes much further. > > Instead of only chatting, you can start with tasks that are built to get work done: > > * Write Doc > * Create Slides > * Create Image > * Build Workflow > > Each task guides your inputs so you get the right result from the start. > > And when things get more complex, you switch to the Superagent. > > ## **Create on-brand Slides & Docs with AI** > > ![D55fe33a Be34 4d02 8c82 65c3787181ab](https://mintcdn.com/nuwacomgmbh/msQ67eH024EqkZjX/images/d55fe33a-be34-4d02-8c82-65c3787181ab-1.png?fit=max\&auto=format\&n=msQ67eH024EqkZjX\&q=85\&s=61a59b7195701d6ecb13b758c7d16b36) > > Create presentations and documents that match your company style — directly inside nuwacom. > > Use built-in templates or upload your own to define how content should look and feel. nuwacom captures key elements like logo, colors, layouts, and structure, and applies them automatically. > > Set voice, audience, and language before generating, then refine everything in a fully integrated editor. When ready, export to PowerPoint or PDF in one click. > > 👉 Content stays consistent and aligned from the first draft. > > ## **Superagent — when work goes beyond a single task** > > The Superagent is a special mode you can activate — similar to choosing a model — when tasks become more complex. > > It uses skills, coordinates agents and tools, pulls in the right data, and executes work step by step across docs, slides, visuals, and workflows. > > For example: > > * A “Morning Briefing” every day at 7:30: pulling your calendar, Jira tickets, and Slack mentions into a clear, prioritized summary > * Preparing you for every customer meeting by gathering the latest details from meeting notes, emails, and your knowledge base > * Creating a full campaign launch package for nuwacom 2.0: including slides, documents, social and visuals > > 👉 From chat, to tasks, to full workflows — work gets done end-to-end. > > ## **Workflow Builder — automate tasks** > > Automate tasks with a clear, structured approach directly inside nuwacom. > > Start with a simple description of what you want to get done. nuwacom turns it into a multi-step workflow that connects actions, logic, and systems. > > Refine each step in a visual editor and schedule workflows to run automatically. > > 👉 Tasks run on their own instead of being repeated manually. > > To activate the new Workflow Builder, please reach out to your company’s nuwacom contact or your Customer Success Manager at nuwacom. > > ## **Notes — record meetings your way** > > Capture meeting content in one place with or without a bot. > > Record meetings directly, use the Meeting Bot, or add manual notes. Transcripts and summaries are generated automatically and combined with your own input. > > Please ensure consent is obtained before automatically capturing meeting notes. > > 👉 One simple flow for all meeting notes. > > ## **Images & Videos — create visuals with direction** > > ![317013da 9eb5 4a07 A4ad 0c57ab0bb0ef](https://mintcdn.com/nuwacomgmbh/msQ67eH024EqkZjX/images/317013da-9eb5-4a07-a4ad-0c57ab0bb0ef-1.png?fit=max\&auto=format\&n=msQ67eH024EqkZjX\&q=85\&s=35b91226e66ade7567a84e1f4e9e7fb6) > > Create images and videos with clear guidance, not just prompts. > > Define format, style, and use case upfront. nuwacom helps you generate visuals that match your intent without needing to fine-tune every detail manually. > > 👉 Visual content that fits your use case from the start. > > ## **Enterprise Search — one place for all your knowledge** > > ![85da44d9 2826 4c65 B268 Bbfe4cfdb0a8](https://mintcdn.com/nuwacomgmbh/msQ67eH024EqkZjX/images/85da44d9-2826-4c65-b268-bbfe4cfdb0a8-1.png?fit=max\&auto=format\&n=msQ67eH024EqkZjX\&q=85\&s=317362001562cb34c5b5da1affdf9cd4) > > Keep your knowledge in sync and access everything in one flow. > > nuwacom brings your internal knowledge into a central knowledge base, while also letting you search across connected apps like Slack, Microsoft Teams, Gmail, Outlook, JIRA, Notion and others > > All relevant information is available in one place and can be used directly in your work. > > 👉 Your knowledge stays connected, accessible, and ready to use. > > ## **A workspace that adapts to you** > > nuwacom 2.0 is a centralized AI workspace that knows your company and gets work done. You can fully customize your experience > > * Pin apps, agents, and workflows to fit your needs > * Access recent work across docs, slides, notes, and tasks in one place > * Refine content directly with chat and inline editing > * Use quick actions like summarizing or translating without breaking your flow > > 👉 A workspace shaped around your work — not the other way around. > > *** > > **New Update – Feb 12, 2026** > > ## **MCP Connectors – Unlock deeper actions for advanced workflows** > > MCP Servers take connectors in nuwacom to the next level. > > Instead of being limited to predefined app actions, admins can now connect MCP Servers to enable a much broader range of system-level actions. This is especially useful for advanced workflows that go beyond standard app integrations. > > Once connected, MCP Server actions appear in Agents and Workflows just like regular app actions. > > ### What’s new > > * Connect MCP Servers directly within your **Workspace Settings** > * Select from common servers or add custom ones > * Use MCP actions inside Agents and Workflows > * Built for admins handling advanced setups > > ### Why it matters > > * **Stronger automations**\ > MCP Servers expose a richer set of actions than typical app connectors. This allows more advanced automation and logic. > * **Better support for complex systems**\ > Some platforms require deeper system access. MCP Servers make this possible inside nuwacom.. > * **Open ecosystem**\ > You are no longer dependent on dedicated app integrations. MCP Servers make it easier to connect new services as your stack grows. > > ### Who is this for? > > This feature is designed for admins and experienced teams who want to build more advanced, action-driven agents and workflows across systems. > > *** > > **New Update – Feb 05, 2026** > > ## **Self-service connectors in your knowledge base** > > Connect data from external apps in seconds. Content is synchronized automatically. > > Folders and individual files from OneDrive, Google Drive, and SharePoint can now be added to the knowledge base on a self-service basis. All data is kept in sync automatically. When files change in OneDrive, Google Drive, or SharePoint, they are updated in nuwacom as well. > > ![Knowledgebaseexternalconnectorsen](https://mintcdn.com/nuwacomgmbh/xig2lzUmPZxuu2Kt/images/knowledgebaseexternalconnectorsen.png?fit=max\&auto=format\&n=xig2lzUmPZxuu2Kt\&q=85\&s=1c41e6054e1f5021cef48df31635d975) > > ### What’s new > > * Add folders and files from external apps to your Knowledge Folders. > * Link individual files from external apps in your personal “My Files” area. > * All connected data is automatically updated when edited externally and stays synchronized across chats, agents, and projects. > > [Learn more here. ](/product/knowledge#sync-external-connectors) > > ## **Model-agnostic document creation** > > Create and edit documents regardless of the selected AI model. Both document creation and editing are now supported by all models, including Claude Opus 4.5, Gemini 2.5 Pro, and GPT-5. > > [Learn more here. ](/product/chat#create-documents) > > ![Mein GIF](https://nuwacom.com/wp-content/uploads/2026/02/documentmodelagnostic.gif) > > ## **Small improvements, big impact** > > * **New model: Claude Opus 4.6**\ > It is Anthropic’s latest large model. It has been further optimized for demanding tasks, especially multi-step analysis, working with documents, tables, and presentations, as well as complex business workflows. > * **Image generation: two new models**\ > With “Imagen 4” and “Imagen 4 Fast”, you now have two new options for image generation. Imagen 4 delivers highly realistic, detailed results and reliably follows your prompts. Imagen 4 Fast offers comparable quality but is significantly faster and more cost-efficient, making it ideal for quick drafts and iterations. > * **Improved document view**\ > Edit your documents with more focus and speed. The most important editor tools are always within reach directly above the document, while ownership, status, and tags can be updated in the top right with just a few clicks. > > *** > > **New Update – Jan 21, 2026** > > ## **User Groups – Share content with teams in seconds** > > With User Groups, you can significantly speed up sharing in nuwacom.\ > Share content instantly by giving entire groups access at once—no need to add users one by one. This way sharing becomes effortless across chats, projects, knowledge folders, and more. > > **Your benefits at a glance** > > * Share content with a single group selection instead of individual users > * Complete recurring sharing tasks in seconds > * Update the group once to update access everywhere. > > ![Usermanagement2](https://mintcdn.com/nuwacomgmbh/tgolVGcUQjuUn1QL/images/usermanagement2-2.gif?s=3cc774b33c57ef56866c6e3c696e536b) > > **How it works** > > 1. Open Workspace-Settings **(Admin rights required)***.* > 2. In the left menu, navigate to *User Management → User Groups.* > 3. Click *Create Group* in the top right corner. > 4. Enter a User Group Name. > 5. Optional: Upload an icon or avatar for the group. > 6. Click *Next*. > 7. Use the Search to add users by clicking. > 8. Click *Create Group* to create the group. > > **Result:** The user group is now created and ready to be used for sharing chats, projects, knowledge folders and more. > > ## **Improved: Active source now visible in the chat** > > ![Source Web En (1)](https://mintcdn.com/nuwacomgmbh/tgolVGcUQjuUn1QL/images/source_web_en\(1\).png?fit=max\&auto=format\&n=tgolVGcUQjuUn1QL\&q=85\&s=4c20a3b26a8313fe8edc0f7f2240f35f) > > Now, whenever you select a source, it’s clearly shown in the chat input—so you always know where the AI will search. You can easily remove the source at any time by hovering over it and clicking the close icon. > > This makes the selected source visible at a glance: > > * **Web** – search the public web. > * **Knowledge** – use your company knowledge base. > * **Web + Knowledge** – combine both sources. > > **How it works** > > 1. Open a new chat. > 2. Click + and activate a source. > 3. Select the source: Web, Knowledge, or both. > 4. Check the input field: The active source is shown directly in the chat input. Remove the source at any time by hovering over it and clicking the close icon. > 5. Enter your prompt and send it. > > **Result:** You always see where the AI will search before you send your prompt — making source management easier and more transparent. # Search Source: https://docs.nuwacom.ai/en/product/search ## Introduction to Search Search brings all content from across nuwacom together in one central place: chats, files, agents, projects, apps, and images appear together, without opening each section individually. Recently edited items are available immediately, can be checked in a preview, and brought straight into chat. ## Key Benefits * **Search across nuwacom**: Chats, files, agents, projects, apps, images, and more appear in one central search, rather than section by section. * **Faster restarts**: Recently used items reappear on open, so work resumes before a new search begins. * **Multiple files in one chat**: Several files are selected and opened together in a new chat as context. ## When to Use Search applies whenever an item needs to be located across the workspace without browsing section by section, or when material is being assembled before a conversation begins. Typical situations include: * Finding a specific chat, file, agent, project, app, or image when its exact location is no longer certain, including files synced to the Knowledge Base through connectors. * Resuming recent work, where recently opened and recently updated items appear before any search term is entered. * Narrowing a large result set to a single owner, project, time period, entity type, or available action through category tabs and filter tokens. * Gathering several files from different categories and opening them together in a new chat as context, rather than attaching them one at a time. ## How It Works ### Open Search 1. Select the magnifying glass icon next to the workspace name in the top-left corner. 2. Use the keyboard shortcut *⌘K* on Mac or *Ctrl+K* on Windows. Either method opens the Search window on top of the current view. ### Search the Workspace Before any text is entered, Search displays recently opened and recently updated items. Full-text search returns every item that contains the entered words. Results are presented across entity types, including chats, files, agents, projects, apps, and images, as well as files synced to the Knowledge Base via connectors. ### Filter Results Category tabs narrow results to a single entity type. Pressing the forward slash (*/*) key opens the filter picker, which offers five filters. Selecting a filter adds it to the search field as a token, for example *owner:jane.doe*. 1. Owner (*owner:*) restricts results to a chosen person. 2. Date (*since:*) restricts results to a time period. 3. Project (*project:*) restricts results to a chosen project. 4. Type (*type:*) restricts results to a chosen entity type. 5. Actions (*actions:*) restrict results to items that support a specific action, such as sharing or archiving. Filters and category tabs combine freely. For example, the Projects tab combined with *owner:jane.doe* returns only Jane's items within that project. ### Preview an Item Hovering over a result or using the *right-arrow key* opens a side panel with details such as the item's owner, an excerpt, and its project or folder location. The preview confirms the correct item before it is opened. Files synced from connected services such as SharePoint or Google Drive open in the source service in a new tab. ## Bring Items into Chat 1. Hovering over a result reveals a checkbox; selecting it chooses the item. 2. A checkmark marks each selected item, including across categories. 3. *Clear selection* removes the selection; *Start chat* opens the selected items as context in a new chat. ## Run an Action on a Result The *Actions* menu, opened with the *right-arrow key*, lists the actions available for the highlighted result, each with a numeric shortcut: 1. *Open* (*1*) opens the item in the current view. 2. *Open in new tab* (*2*, *⌘⇧*) opens the item in a separate browser tab. 3. *Start chat* (*3*) opens the item in a new chat as context. 4. *Add to selection* (*4*, *⌘*) adds the item to the current multi-item selection. The available actions reflect the type of the highlighted result. ## FAQ > **How is Search opened?** > > Through the magnifying glass icon next to the workspace name in the top-left corner, or with *⌘K* on Mac or *Ctrl+K* on Windows. > **Which types of items appear in results?** > > Chats, files, agents, projects, apps, and images. Full-text search returns whatever is entered. > **Why are some categories not visible?** > > The available categories reflect the features and permissions in a given workspace. Items and entity types without access do not appear. > **Can results be narrowed to a specific owner or time period?** > > Yes. The */* key opens the filter picker, which offers *owner:*, *since:*, *project:*, *type:*, and *actions:*, each added to the search field as a token. Category tabs narrow results to a single entity type and combine with filters. > **What happens to files from connected services?** > > Files synced from services such as SharePoint or Google Drive open in the source service in a new tab. > **Which actions are available on a result?** > > The *Actions* menu lists *Open* (*1*), *Open in new tab* (*2*, *⌘⇧*), *Start chat* (*3*), and *Add to selection* (*4*, *⌘*). The available actions reflect the type of the highlighted result. > **How are several items added to a chat at once?** > > Hovering over a result reveals a checkbox, which selects items including across categories. *Start chat* opens the selected items together as context in a new chat. # Workflows Source: https://docs.nuwacom.ai/en/product/workflows ## Introduction to Workflows Workflows in nuwacom automate recurring processes. Required steps are defined once, then executed in sequence. Agents, logic, integrations, and external tools can run within one workflow. A workflow can be described in chat. nuwacom then generates a multi-step workflow that connects actions, logic, and external systems. The visual editor keeps every step adjustable, extendable, and transparent. Suggested prompts provide a starting point for common workflow scenarios. They help explore recurring use cases and turn them into editable workflow structures. ## Key benefits * Automate recurring processes with defined steps. * Combine agents, logic, integrations, and external tools in one workflow. * Reuse results from earlier workflow steps in later steps. * Add manual approvals before a workflow continues. * Create reusable outputs such as documents, email drafts, designed documents, and slide decks. ## When to use workflows Workflows are suitable when tasks repeat regularly, follow a clear structure, or require several systems to work together. Typical use cases include creating and reviewing reports, monitoring incoming data, sending automated emails, processing customer inputs, conducting research, and generating summaries. Use tasks for ad hoc work in chat. Use workflows for recurring, automated processes with fixed steps. ## Creating Workflows ### Create the First Version in Chat ![Createworkflowen](https://mintcdn.com/nuwacomgmbh/iv1Cu73TG30JH6P1/images/createworkflowen.png?fit=max\&auto=format\&n=iv1Cu73TG30JH6P1\&q=85\&s=c72c2fd9bb3af4b2166c2221b8def5d0) The chat generates a complete, multi-step workflow from a plain-language description. 1. Open the chat and select the task *Build workflow* or click *+ New Workflow* in the workflow module. 2. Describe what the workflow should do, including the goal, the information to process, and the final action. 3. Send the prompt. Example prompt: *"Create a workflow that helps me plan my daily tasks. Analyze my calendar and suggest suitable time slots based on availability and deadlines."* From this description, nuwacom assembles the multi-step workflow automatically. ### Review the Workflow in the Workflow Builder The Workflow Builder shows the generated structure on the canvas, where each component and connection can be verified before production use. 1. Open the workflow in the Workflow Builder. 2. Review the workflow structure on the canvas. 3. Check how the components are connected. 4. Review where logic, agents, or integrations are used. 5. Adjust missing steps or existing configurations. 6. Test the workflow before production use. Testing verifies that all steps run correctly and that the results match the expected outcome. > [!NOTE] > > The prompt *“Help me fix an error”* starts an AI-supported workflow check. nuwacom analyzes the workflow, identifies issues, and shows how to resolve them. ### Run the Workflow The *Run Progress* panel tracks execution step by step and surfaces the status of each one. 1. Select *Run* in the upper-right corner. 2. Follow the execution in the *Run Progress* panel and check the status of each step: * Green: completed * Purple: running * Grey: not started * Red: failed or cancelled 3. Expand a step to view details or error messages. 4. Review the results and confirm that all steps and outputs are correct. Once the workflow has been successfully tested and reviewed, publishing the version makes it available for production use. ### Publish a Workflow The *Publish* button in the top bar sets a draft workflow live. Once published, the workflow becomes active and runs automatically whenever its trigger fires, and its status changes to *Live*, indicated by a green dot in the top bar. The *Publish* button appears disabled while the workflow is already live. ### Unpublish a Workflow *Unpublish* is available in the three-dot menu (⋮) in the top bar. Unpublishing returns a live workflow to draft state, so it no longer runs when its trigger fires. The workflow and its configuration remain saved and editable, and it can be published again at any time. ### Execution History The three-dot menu (⋮) also provides the *Execution History* of a workflow. It opens a panel on the left listing every past run of the workflow in chronological order, with the most recent at the top. Each entry shows a status icon, a timestamp, and the run duration. Runs can be filtered by status using the *All*, *Running*, *Completed*, *Failed*, and *Cancelled* tabs. Selecting an entry displays that run on the canvas on the right, showing each step with its individual status and processing time. This makes it possible to trace how a specific run progressed and to locate where a failed run stopped. ### Export and Import a Workflow The three-dot menu (⋮) also provides *Export JSON* and *Import JSON*. *Export JSON* downloads the workflow's complete configuration as a JSON file, and *Import JSON* loads a workflow from a JSON file back into the workspace. This is useful whenever a workflow needs to move between people or environments. As of today, workflows cannot be shared with other users directly. Exporting the JSON file and handing it over is the way to pass a workflow to someone else, who can then import it into their own workspace. The same approach works for keeping a backup before making major changes, duplicating a workflow as a starting point for a new one, or moving a workflow from a test to a production workspace. *** ## Managing Workflows The metrics bar at the top of the Workflows page summarizes overall activity across the last 30 days, including the total and published workflow counts, total runs, success rate, and failed runs. The page then presents every workflow in the workspace in a single table. Each row represents one workflow, and the columns describe its key attributes and recent activity. ### Metrics Bar A metrics bar at the top of the page summarizes activity across all workflows: * **Workflows:** Total number of workflows, with a count of how many are published. * **Runs:** Number of runs recorded in the last 30 days. * **Success rate:** Share of successful runs over the last 30 days. A dash appears when no runs exist yet. * **Failed:** Number of failed runs in the last 30 days. Hovering over the graphic under Runs or Failed displays a tooltip with the respective date and the number of runs. The metrics cover the last 30 days. ### Create a new Workflow The *+ New Workflow* button in the top-right corner opens a new Workflow in draft state, which stays a draft until it is completed and published. The new workflow opens with two ways to build it: the Copilot panel on the left builds the workflow from a plain-language description of what it should do, while the editor on the right starts from scratch. ### The Table View Every workflow in the workspace appears as a single row, and each column describes one of its properties. A three-dot button at the end of each row opens that workflow's actions. | Column | What it shows | | :- | :- | | Name | The workflow title, used to identify and open it | | Status | Whether the workflow is *Live* or *Draft* | | Trigger | What starts the workflow, such as *Scheduled, Manual, Form, Webhook, App,* or *nuwacom* | | Steps | A preview of the actions the workflow performs, shown as icons in run order | | Runs (30 d) | The number of runs recorded in the last 30 days | | Success | The result of the most recent runs, or a dash when no run has occurred | | Last Run | The result and relative time of the most recent run, e.g. "Failed · 4 months ago" | | Last modified | When the workflow was last edited | | Created | When the workflow was first created (optional) | The sort control above the table reorders the rows, and the columns panel adjusts which columns are shown. ### Workflow Actions The three-dot button at the end of each row opens the actions for that individual workflow: * **Pin to sidebar:** Adds the workflow to the sidebar for fast, recurring access. * **Edit:** Opens the workflow to adjust its trigger, steps, and details. * **Rename:** Changes the workflow name in place. * **Delete:** Removes the workflow permanently. This action appears in red to mark it as destructive. ### Workflow Status The Status column shows whether a workflow is running: * **Draft:** The workflow is saved and editable but does not run automatically. * **Live:** The workflow is active and runs whenever its trigger fires. The published count in the metrics bar reflects how many workflows are currently active. ### Triggers Triggers start a workflow. A workflow starts manually or automatically when a defined event occurs. The Trigger column shows which type applies to each workflow: * **Scheduled Trigger:** Runs the workflow on a schedule, at defined times or intervals. * **Manual Trigger:** Starts the workflow by hand, with custom inputs. * **Form Trigger:** Starts the workflow when a form is submitted. * **Webhook:** Starts the workflow via an incoming HTTP webhook. * **App Trigger:** Starts the workflow from an event in a connected app. * **nuwacom Trigger:** Starts the workflow from an event in the nuwacom workspace. ### Steps Steps are the individual actions a workflow performs. The Steps column previews them as a row of icons that run in order. Steps are built from nodes, logic controls, agents, integrations, and output nodes. ### Run History and Results Three columns track how each workflow has performed: * **Runs (30 d):** Number of times the workflow ran in the last 30 days. * **Success:** Result of the most recent runs, shown as Succeeded, Failed, Cancelled, or a dash when no run has occurred. * **Last Run:** Result and relative time of the most recent run, for example "Failed · 4 months ago". The Last modified column shows when the workflow was last edited, which distinguishes recently updated automations from older ones. ### Managing Columns The columns control next to Filter opens a panel for showing or hiding table columns. A green toggle keeps a column visible, and a gray toggle hides it. The table updates immediately. Shown by default: Name, Status, Trigger, Steps, Runs, Success rate, Last Run, and Last modified. The Created column is available but hidden until its toggle is enabled. ## Filtering, Searching, and Sorting Workflows Three controls narrow down the workflow list. A search term and a sort order each apply on top of the active filter. ### Search The search field filters the list by workflow name as characters are entered. ### Filter The Filter control opens a dropdown that narrows the list by status. A checkmark marks the option currently applied, and a count next to each option shows how many workflows match: * **All:** Shows every workflow regardless of status. This is the default view. * **Live:** Shows only active workflows, marked with a green dot in the Status column. * **Drafts:** Shows only saved, editable workflows that are not yet live and do not run automatically. Selecting an option immediately filters the list to the matching workflows. The counts update to reflect the current workspace and indicate at a glance how many workflows are live versus still in draft. ### Sort The sort control on the right opens a dropdown of ordering options. A checkmark marks the option currently applied. By default the list is sorted by Recently Updated. * **Recently Updated:** Orders the list by the last time each workflow was edited, most recent first. * **Recently Run:** Orders the list by the time of the most recent run, most recent first. * **Most Runs:** Orders the list by number of runs, from the most-run workflow down. * **Recently Created:** Orders the list by creation date, newest first. * **A–Z:** Orders the list alphabetically by workflow name. * **Schedule:** Orders the list by the workflows' scheduled run times. Selecting an option reorders the list immediately and applies on top of the active filter and search term. *** ## Components of a workflow ### Triggers Triggers start a workflow. A workflow can start manually or automatically when a defined event occurs. Common trigger types include manual triggers, form submissions, and scheduled triggers. Scheduled workflows can create daily reports, weekly meeting summaries, or monthly data analyses. ### Nodes Nodes are the individual steps within a workflow. Each node performs one specific action. Typical node actions include running an agent, processing data, transforming data, calling an external API, sending an email, and applying logical conditions. Nodes can be connected to create simple automations or complex multi-step workflows. ### Logic controls Logic controls define how workflow steps are executed. They allow the workflow to adapt to incoming data. Supported controls include If/Else conditions, loops, and parallel steps. If/Else conditions define the next path. Loops repeat actions for multiple items. Parallel steps run tasks at the same time. ### Agents Agents can run as workflow steps. They can generate content, analyze data, summarize documents, perform research, and make decisions. Agent results can be passed to other agents. This enables multi-agent automations within one workflow. Example workflow: one agent creates a job description. A second agent reviews and refines it. The final result is sent by email. ### Integrations Integrations connect workflows with external tools and services. Examples include email services, external APIs through HTTP requests, and connected data systems. This allows workflows to move data between systems automatically. ### Output Workflow Output Nodes create reusable nuwacom content at the end of a workflow, including documents, email drafts, and slides. Outputs can be saved to the Knowledge Base, and optionally attach the result to a project. *** ## Review and monitor workflows Manual approval steps pause a workflow until a result is reviewed, approved, or rejected. The workflow view provides an overview of the current execution status and supports detailed analysis of individual runs. The workflow view also allows you to: * Inspect results for individual steps. * Identify and analyze errors. * Review previous workflow runs. * Verify how manual approval decisions affect the workflow. *** ## nuwacom Trigger ![Workflowsnuwacomtrigger](https://mintcdn.com/nuwacomgmbh/iv1Cu73TG30JH6P1/images/workflowsnuwacomtrigger.png?fit=max\&auto=format\&n=iv1Cu73TG30JH6P1\&q=85\&s=2bb4bec3c25938dfa95cae941f662fc8) nuwacom-native workflow triggers start published workflows automatically when defined events occur inside nuwacom. They remove the need for manual launches, external webhooks, or third-party integrations. Examples include creating a meeting note or completing the processing of a Knowledge Base file. Trigger data becomes available to later workflow steps and AI agents. ### Key benefits of nuwacom-native workflow triggers * Start workflows as soon as the configured event occurs. * Limit workflow runs to relevant events through filters. * Provide agents with the data that triggered the workflow. * Configure internal events without external connections. * Display available trigger fields in the variable picker. ### Supported use cases * Start a published workflow when a new meeting note is created. * Start a workflow when a Knowledge Base file finishes processing. * Restrict a trigger to files added to a specific Knowledge Base folder. * Pass event details to downstream workflow steps and AI agents. * Reference trigger data through the built-in variable picker. ### Example A workflow can start automatically when a file finishes processing in a selected Knowledge Base folder. The workflow can then pass the file details to an AI agent for classification, summarization, or further processing. *** ## nuwacom Outputs ![Workflowsnuwacomoutputs](https://mintcdn.com/nuwacomgmbh/iv1Cu73TG30JH6P1/images/workflowsnuwacomoutputs.png?fit=max\&auto=format\&n=iv1Cu73TG30JH6P1\&q=85\&s=d538cfc68059cc328d5b23f1c7c4be87) Workflow Output Nodes create reusable nuwacom content automatically. A workflow can generate documents, email drafts, designed documents, or slide decks instead of ending with a text response. Generated content can be saved directly to the Knowledge Base and attached to a project. Workflow results remain available for later work and later workflow steps. ### Key benefits of Workflow Output Nodes * Create finished deliverables directly from workflow results. * Reduce manual copying between tools. * Apply consistent writing, formatting, and brand guidelines. * Save outputs in the correct Knowledge Base folder or project. * Reuse generated content in later workflow steps. * Prevent unnecessary duplicates when rerunning an output step. ### What Workflow Output Nodes can create Workflow Output Nodes can create HTML documents, plain text documents, email drafts, slide decks, and designed documents using selected skills and bundled assets. The node can generate new content or save the output from an earlier step unchanged. ### When to use Workflow Output Nodes Workflow Output Nodes are suitable for recurring processes that produce reusable content. Relevant scenarios include summaries from structured workflow results, recurring documents, email drafts for review, slide decks from collected information, branded documents from standardized instructions, and project documentation. ### Example A workflow receives meeting notes from an earlier step. A Workflow Output Node turns the result into a structured HTML document. *** ## FAQ > **When should workflows be used?** > > Workflows fit recurring tasks with a clear structure. They also support processes that connect multiple systems. > **What is the difference between tasks and workflows?** > > Tasks support ad hoc work in chat. Workflows automate recurring processes with fixed steps that run again and again. > **Does every workflow need to be built manually?** > > No. A workflow can be described in chat. nuwacom generates a first version that can be adjusted in the Workflow Builder. > **Can a workflow run automatically?** > > Yes. Workflows can run on schedules, from form submissions, or from internal nuwacom events. The selected trigger defines how the workflow starts. > **Can workflows use more than one agent?** > > Yes. A workflow can include multiple agent steps. The output of one agent can be passed to the next. > **Can results be reviewed before a workflow continues?** > > Yes. Human-in-the-loop approval steps allow results to be reviewed, approved, or rejected before the next step runs. > **Can a workflow create documents or saved outputs?** > > Yes. Output nodes can create documents, email drafts, designed documents, and slide decks directly in nuwacom. > **How can I share a workflow?** > > Workflows cannot currently be shared directly through a sharing button. To share a workflow, download it as a JSON file via the three dot menu and send the file to another user. > > The recipient can upload the JSON file to import the workflow into their own workspace. # Research Source: https://docs.nuwacom.ai/en/research With the “Research” task, you can directly access web search, your internal knowledge, or connected apps within the chat and receive structured, up-to-date results in seconds – without manual searching, comparing, or compiling. Use suggested prompts to instantly start with common research questions or get inspired, for example: * *Research global technology trends for the next 5 years* * *Analyze emerging AI use cases across different industries* * *Understand workforce and leadership trends for modern organizations* * *Identify sustainability trends companies should prepare for* > [!NOTE] > > In nuwacom, you research using the most relevant sources – such as the web, internal knowledge, or connected apps. Relevant content is automatically analyzed, filtered, and structured. The sources used are displayed transparently, so you can review and verify results. ## **How it works** ![Research](https://mintcdn.com/nuwacomgmbh/tzlUXQ5D4ImeIx2T/images/research.png?fit=max\&auto=format\&n=tzlUXQ5D4ImeIx2T\&q=85\&s=050282f6c3999598f3227cbac237020c) **1. Open chat and start research** Open the chat and start the “Research” task **2. Write your prompt** Formulate a clear and specific request, for example:\ *“Analyze the key competitors of Canyon in the European e-bike market. Compare positioning, strengths and weaknesses, pricing structure, and key differentiators.”* **3. Select source** Choose the appropriate source to define the context, such as web, internal knowledge, or connected apps. By default, the web is selected. **4. Generate results** nuwacom searches the selected sources, analyzes relevant content, and gathers key information about the market, companies, and differences. **5. Review results** You receive a structured competitive analysis with a market overview, company profiles, clear differences between providers, and depending on your request, visualizations such as a positioning matrix. The sources used are also displayed. This allows you to trace information transparently, assess the quality of the results, and jump directly to the original sources if needed. **6. Deepen your analysis (follow-ups)** Build directly on the results and ask follow-up questions, for example:\ “Which of these competitors are particularly strong in the B2B segment?”\ “How do the marketing strategies of the top 3 providers differ?” **7. Reuse results** Use the results directly, for example for: * Market and competitive analysis * Presentations or strategy documents * Decision-making within your team This allows you to move seamlessly from research to execution. ## **Work seamlessly with your results** You can directly reuse your analysis results and integrate them into your workflows. For example, you can: * share the entire chat or individual responses with colleagues to ensure transparency or collaborate further * move relevant content into a project or knowledge folder to structure and expand it * use the response as a basis for an email, adjust it if needed, and send results quickly * create structured documents from the results and continue editing them in the editor, for example for reports or presentations # Measures Source: https://docs.nuwacom.ai/en/security/measures In this article, you will learn more about the security measures we have implemented to protect your data. Our goal is to provide you with a high level of security and privacy so that you can rely on our services. ## **Security and Privacy Measures** | Certifications | ISO 27001:2022 | | :- | :- | | GDPR Compliance | Yes, we are GDPR compliant. For more information, please refer to our DPA. | | Application Hosting Location | We host nuwacom in the Microsoft Azure EU Cloud. The preferred server location is Frankfurt am Main, Germany. | | LLM Hosting Location | The models currently used on the nuwacom platform are primarily hosted in Frankfurt, Germany, and Gävle, Sweden. | | Data Center Certifications | AWS: Visit the [AWS Compliance page](https://aws.amazon.com/en/compliance/) and the [ISO Certification page](https://aws.amazon.com/de/compliance/iso-certified/). Azure: Visit the [Azure Compliance page](https://learn.microsoft.com/en-us/azure/compliance/) and the [ISO Certification page](https://learn.microsoft.com/en-us/azure/compliance/offerings/offering-iso-27001). | | Backups | Full daily backups | | Backup Location | Physically separate Azure location within the European Union | | Availability | 99.99% uptime - Visit our [Status Page](https://status.nuwacom.ai/) in case of issues. | | Authentication | Strong and flexible password policies, multifactor authentication (MFA), and single sign-on (SSO) | | Authorization | [Role-based access control (RBAC)](https://docs.nuwacom.ai/product/usermanagement#roles-and-permissions) | | Data Encryption | TLS 1.2/1.3 for data in transit, AES-256 for data at rest | | IPv6 Support | Yes | ## **Data, Privacy, and Security of LLMs** Your inputs, outputs, and company data are secure and are not used for other customers or to improve LLM models. These data are neither accessible to other customers nor to OpenAI or other providers and are not used to enhance LLM models or optimize Microsoft and third-party products. ## **Client separation** Each client has dedicated hardware for data, separate storage, isolated RDB disks, an individual vector database, a unique nuwacom domain, and independent scaling capabilities. This customized infrastructure ensures optimal performance and security for every client. For further information or any questions, feel free to contact us at any time. # Skill Templates Source: https://docs.nuwacom.ai/en/skill-templates ## What Are Skill Templates Skill Templates are ready-made Skills that nuwacom provides as a browsable collection covering common use cases. Each template ships with a predefined structure, tone, and rule set, so a working Skill is available immediately instead of being written from scratch. Templates are organized by business area, which makes it fast to locate a fit and then adapt it to specific requirements. Templates are best suited for a quick start: a matching template delivers a tested foundation that only needs to be aligned with the target process. Every Skill created from a template remains fully configurable in the code editor after creation. ## Browsing the Template Library The template library opens from the Skills page via the "Browse Templates" button. Templates are organized into categories in the left sidebar: All Templates, HR, Legal, Product, Sales, Support, and Writing. Selecting a category filters the available templates, and the search field locates a specific one. ### Skill Templates The library provides two ways to narrow down the selection: * **Search**: Use the Search templates field to find templates by name or description. * **Categories**: Filter by business area to focus on relevant templates. Available categories include HR, Legal, Product, Sales, Support, and Writing. Representative templates per category: 1. **HR**: HR Policy Handbook, Interview Guidelines, Job Ad Structure & Tone 2. **Legal**: Compliance Review Skill, Legal Clause Library, Privacy Guidelines Skill 3. **Product**: Product Glossary, Product Requirements Skill, Research Synthesis Skill 4. **Sales**: Customer & ICP Profile, Pricing & Packaging, Sales Playbook 5. **Support**: Escalation Policy, Support Response 6. **Writing**: Brand Kit Guardian, Tone of Voice, Writing Style Guide ## Template Details Selecting a template card opens the Create a New Skill preview. It displays the template name, a Skill icon, and a natural-language prompt field for describing changes, along with the full configuration under two building blocks: Description and Instructions. The Create Skill button adds a copy of this configuration to the workspace. The HR Policy Handbook Skill template illustrates how these building blocks fit together. ### Description The Description defines what the Skill covers and when it applies. In the HR Policy Handbook Skill, it states that the Skill holds HR rules, policies, benefits, absences, onboarding, and employee communications, and lists the inquiries that activate it, even when HR is not named explicitly. ### Instructions The Instructions define how the Skill retrieves, interprets, and communicates its content. In the HR Policy Handbook Skill, they are structured into named sections: * **Purpose**: Grounds every response strictly in official documentation, with no invented policy details. * **KB-Grounding**: Sets the retrieval behavior, including targeted queries, fallback synonyms, and the rule not to answer without relevant content. * **Tool Rules**: Maps each situation to the correct tool and excludes non-authoritative sources such as Web Search. * **Process**: Breaks the task into numbered steps, from request to response, gap handling, and next step. * **Scope & Boundaries**: Separates in-scope tasks from situations that are always escalated. * **Output Format**: Provides ready-made response structures for policy answers, onboarding, and communications. * **Tone & Style**: Sets the language register as clear, plain, professional, and consistent with source terminology. This structure shows the recommended pattern: a purpose that frames the Skill's mandate, followed by explicit, rule-based instructions that govern retrieval, tool usage, scope, and output. ## How to Create a Skill from a Template 1. Open "Browse Templates" on the Skills page. 2. Select a category and choose a template, for example HR Policy Handbook, then review its description and instructions in the preview. 3. Customize the copy in the code editor. The SKILL.md editor refines the instructions, the description field defines what the Skill covers, and the files panel manages assets such as reference documents, logos, or fonts. 4. Select *Save*. A copy of the template is added to the Skill library, for example "Copy of HR Policy Handbook", and is private by default. The created Skill is fully editable and independent of the original template. Changes to it do not affect the template or any other Skill created from it. > ***Tip***\*: A copy inherits the full template configuration, so the fastest path to a working Skill is to create it from the closest match and refine only the instructions and assets that differ from the target process.\* ## Customizing Your Copy Once created, the Skill is fully configurable in the code editor. Each element maps to what can be changed: * **Adjust instructions**: The code editor refines the instructions in structure, tone, or rules. * **Add new file**: Individual files such as reference documents, logos, or fonts are attached directly to the Skill. * **Add new folder**: Folders keep bundled assets organized within the Skill. * **Upload file**: Assets from the local system are uploaded into the Skill. * **Share, activate, and download**: The Skill can be shared with the team to standardize outputs, set active via the toggle to make it available in chat, and exported as a file for backup or reuse elsewhere. ## FAQ > **Does creating a Skill from a template change the original?** > > No. The copy is fully independent, so changes never affect the template or other Skills made from it. It appears as a copy, for example *Copy of HR Policy Handbook*, and is private by default. > **Can a Skill from a template be edited afterwards?** > > Yes. It stays fully configurable in the code editor: the *SKILL.md* editor refines the instructions, the description field defines what the Skill covers, and the files panel manages assets. > **How is the right template found quickly?** > > Search finds templates by name or description, and the category sidebar filters by business area — HR, Legal, Product, Sales, Support, and Writing. # Slides Source: https://docs.nuwacom.ai/en/slides Create slides directly in nuwacom – without starting from scratch. Use templates, suggested prompts, as well as language, tone of voice, and target audience to tailor your presentation. Simply describe your topic, and nuwacom automatically generates structured slides with relevant content and layouts. The presentation opens in the editor, where you can edit content via chat or inline editing and present it instantly. ## **How to create slides** ![Promptslides](https://mintcdn.com/nuwacomgmbh/tzlUXQ5D4ImeIx2T/images/promptslides.png?fit=max\&auto=format\&n=tzlUXQ5D4ImeIx2T\&q=85\&s=d83e769aaf1fdf956c88a73de408207f) **1. Open chat** Open the chat. **2. Select task** Click on “Create slides”. **3. Describe slides** Describe the presentation you want to create in the input field.\ Example: *“Create a presentation for the launch of our new e-bike model 'Gravel Bike' including target audience, USPs, market potential, and marketing strategy.”* **4. Define language, tone, and audience** Optionally select language, tone of voice, and target audience to align the content. **5. Upload or choose template** Upload your own templates or choose one to create your presentation in your corporate design. **6. Start generation** Start the generation. **7. Edit slides** Edit your presentation directly in the editor – via chat or inline editing. **8. Present slides** Present your slides directly from nuwacom without leaving the platform. > [!NOTE] > > The AI uses your template as a reference. It recognizes your corporate design (colors, fonts, overall look), but builds the slides dynamically. > > This means: > > * Layout and structure are automatically adapted to the content and may differ from the master > * Text is repositioned and reflowed to ensure good readability > * Color rendering may vary slightly > > When using AI, small to moderate deviations are normal. That’s why we enable you to edit the slides in the editor or via chat. There you can: > > * Adjust or restructure layouts > * Replace or add images > * Insert or update logos and icons > * Edit colors, spacing, and positioning > > 👉 Result: High-quality slides that align with your corporate design. ## **Supported template files** You can use the following file formats as templates:\ .ppt, .pptx, .png, .jpg, .jpeg, .html **Note**\ The maximum file size per upload is 50 MB. ## **Edit slides** ![Editslides (1)](https://mintcdn.com/nuwacomgmbh/tzlUXQ5D4ImeIx2T/images/editslides\(1\).png?fit=max\&auto=format\&n=tzlUXQ5D4ImeIx2T\&q=85\&s=e66182bc3a32171406023bc714a7b57c) #### Edit and extend content flexibly You can edit and extend your presentation at any time – either via chat or directly with inline editing in the editor. Adapt content flexibly, add new information, or refine existing slides directly in the desired format. * Add or create new content * Enhance slides with infographics * Rephrase content * Make visual adjustments * Edit text, structure, and formatting directly #### **Add or generate images** You can enhance your slides visually by uploading images or generating new ones. ![Editimageinslides](https://mintcdn.com/nuwacomgmbh/tzlUXQ5D4ImeIx2T/images/editimageinslides.png?fit=max\&auto=format\&n=tzlUXQ5D4ImeIx2T\&q=85\&s=b986870c2b6068844f6fa6db918c8372) 1. Click on the image icon on the slide. 2. Choose one of the options: * Upload image * Generate image * Select image from gallery > [!NOTE] > > When generating, simply describe the image you want.\ > Example: “High-end e-gravel bike with carbon frame, integrated battery, matte Tangelo Orange finish, clean lines, urban evening light.”\_The AI will generate suitable visuals for your presentation. ## **Present, share, or reuse slides** ![Presentslides](https://mintcdn.com/nuwacomgmbh/tzlUXQ5D4ImeIx2T/images/presentslides.png?fit=max\&auto=format\&n=tzlUXQ5D4ImeIx2T\&q=85\&s=d1fb0169751366afd4605837a5bffe18) Your presentation can be presented or reused directly from nuwacom – without leaving your workflow. You can: * present the presentation directly from nuwacom * share the presentation * move it to a project or knowledge folder * download it as .PPTX or .PDF This keeps your presentation centrally organized and allows you to present it or further develop it with your team. ## **Organize and manage slides** Your created slides are automatically saved and can be accessed anytime via *AI Apps > Slides*. You can further organize your presentation and prepare it for collaboration. The following functions are available: * **Save version** – save a new version of your presentation. * **View version history** – access previous versions and restore them if needed. * **Add owners** – assign people responsible for the presentation. * **Add tags** – structure your presentations using keywords. These features help you manage presentations in a structured way and collaborate effectively with your team. # SSO & User Sync Source: https://docs.nuwacom.ai/en/sso-and-usersync With Single Sign-On (SSO), you sign your team in to nuwacom through your identity provider. In addition, you can automatically sync roles, groups, and user groups through user synchronization. You'll find these settings under **Workspace Settings > SSO & User Sync.** > [!NOTE] > > **Note:** SSO & User Sync is currently only available in multi-tenant environments, i.e. at app.nuwacom.ai or app-de.nuwacom.ai. This feature is not currently available in single-tenant environments.\ > \ > For single-tenant environments, integration is currently handled through a custom SSO setup. If you'd like to set up SSO for your environment, please contact us at [support@nuwacom.ai](mailto:support@nuwacom.ai). Our team will review your requirements and support you with the next steps. #### Prerequisites To set this up, you need admin rights in nuwacom and access to your identity provider. *** ## Enable Single Sign-On First, enable SSO and specify which email domains should sign in through your identity provider. 1. Open the workspace settings. 2. Go to SSO & User Sync. 3. Turn on the **Enable SSO** toggle. 4. Under **Email Domains**, enter the domain for which SSO should apply. **Example:** ```text theme={null} firma.de ``` Team members with this email domain will then be automatically redirected to your identity provider. 5. Then select the appropriate provider type to connect your identity provider to nuwacom via **OpenID Connect (OIDC)** or **SAML 2.0**. ### OpenID Connect (OIDC) If you select **OpenID Connect (OIDC)**, enter your identity provider's connection details. Enter the following values: | **Field** | **Description** | | :- | :- | | Discovery URL | The discovery URL of your identity provider. | | Client ID | The client ID of the application from your identity provider. | | Client Secret | The client secret of the application from your identity provider. | **Example of a discovery URL:** ```text theme={null} https://login.example.com/.well-known/openid-configuration ``` Then copy the **Redirect URI** shown in nuwacom and enter it when registering the application in your identity provider. Then click **Save Configuration**. ### SAML 2.0 If you select **SAML 2.0**, enter your identity provider's metadata. You can enter the metadata in two ways: | **Option** | **Description** | | :- | :- | | URL | Enter the metadata URL of your identity provider. | | XML | Paste the metadata XML directly into the text field. | Then copy the values shown in nuwacom and enter them in your identity provider: | **Value** | **Description** | | :- | :- | | ACS URL | The URL to which your identity provider sends the SAML response. | | SP Entity ID | The unique identifier of nuwacom as the service provider. | Then click **Save Configuration**. ### Advanced Settings Here you define how external attributes from your identity provider are mapped to user attributes in nuwacom. Open this section and review the attribute mapping. By default, the following mappings are set, for example: | Claim from IdP | User Attribute | | :- | :- | | email | email | | given\_name | firstName | | family\_name | lastName | Click **Add Mapping** if you want to transfer additional attributes. Enter the external claim from your identity provider and map it to the appropriate user attribute in nuwacom. Then select the appropriate sync mode: | **Sync Mode** | **Description** | | :- | :- | | Force | Attributes are overwritten on every login. | | Import | Attributes are only imported on the first login. | | Legacy | Existing sync mode for older configurations. | Then click **Save Configuration**. *** ## Enable SCIM User Sync With SCIM, you automatically sync users from your identity provider to nuwacom. 1. Turn on the **Enable SCIM User Sync** toggle. 2. Copy the displayed **SCIM Endpoint URL** and enter it in your identity provider. 3. Click **Generate Token** to create a bearer token. 4. Copy the bearer token and enter it in your identity provider for the SCIM connection. > [!NOTE] > > **Important:** Treat the bearer token like a password. Do not share it publicly and store it only in secure locations. ### Role Mapping Here you define which IdP roles or IdP groups correspond to which workspace roles in nuwacom. 1. In the **Role Mapping** section, click **Add Rule**. 2. Enter the name of the role or group from your identity provider. **Example:** ```text theme={null} marketing-team ``` Then select which workspace role users with this IdP role or group should receive: | **Role** | **Description** | | :- | :- | | Administrator | Has full access to the workspace and its settings. | | Editor | Can create and edit content. | | Reader | Can read shared content. | If needed, click **Add Fallback Rule** to set a default role for all other users. Then click **Save Rules**. > [!NOTE] > > **Note:** The first matching rule is applied. Users without a matching rule will not get access to the workspace if no fallback rule is defined. ### Sync User Groups If you'd also like to use SCIM groups as user groups in nuwacom, enable **Synchronize SCIM Groups as User Groups**. nuwacom will then automatically create and update user groups based on SCIM group memberships. *** ## Troubleshooting > **SSO login isn't working. What should you check?** > > Make sure the email domain is entered correctly and that SSO is enabled for this domain. Also check that the required values are configured exactly in your identity provider: > > * For OIDC: Redirect URI > * For SAML: ACS URL and SP Entity ID > **Users aren't being redirected to the identity provider. What could be causing this?** > > Check whether the user's email domain is set up in nuwacom. Automatic redirection only works for users whose email domain matches a domain configured for SSO. > **A user doesn't have access to the workspace. What should you check?** > > Check whether there is a matching role mapping for the user. If no mapping rule applies and no fallback rule is defined, the user will not get access to the workspace. > **Users aren't being synced via SCIM. What should you check?** > > Make sure SCIM user sync is enabled. Also check whether the SCIM endpoint URL is entered correctly in your identity provider and whether the bearer token is still valid. > **Groups aren't being imported as user groups in nuwacom. What should you check?** > > Check whether **Synchronize SCIM Groups as User Groups** is enabled. Also make sure that group memberships are correctly maintained in your identity provider and transferred via SCIM. > **Changes from the identity provider don't appear in nuwacom. What should you check?** > > Synchronization is triggered by your identity provider. Check there whether the change was successfully sent to nuwacom. Also make sure that SCIM is active and the connection to nuwacom is working. > **Attributes such as first name, last name, or email aren't imported correctly. What should you check?** > > Open **Advanced Settings** and review the attribute mapping. Make sure the claims from your identity provider are mapped to the correct user attributes in nuwacom. > **Does the SSO configuration apply to the entire organization, or does it need to be set up for each workspace?** > > SSO is configured at the organization level and automatically applies to all workspaces within the organization. A separate setup for each workspace is not required. Changes to the SSO configuration are also applied across the entire organization. # Usage Rules Source: https://docs.nuwacom.ai/en/usage-rules Each nuwacom user license includes AI usage for normal day-to-day work. This covers chat, knowledge search, summaries, meetings, documents, presentations, agents, workflows and apps. The included AI usage allows companies to plan their usage and costs reliably. Additional costs only arise if the company approves further AI usage or upgrades to a higher user profile. ## **How AI Usage Works** Using an AI model creates usage-based model costs. Text, files, and responses are processed in units called tokens. Tokens are small parts of words, full words, or characters. AI usage includes the information sent to the model, the processing performed within the nuwacom platform, and the generated response. This includes, for example: * the current request * attachments and attached company knowledge * relevant parts of the previous chat history * the generated response * individual steps within agents or workflows The selected model also affects usage. More advanced models may use more AI usage for the same task than more cost-efficient models. ## **Context and Usage** Context includes all information the AI model needs to consider when generating a response. The amount of context increases, for example, when a chat includes a large attachment or several files. It may also grow during a longer chat because earlier questions, answers, and files can be included in later requests. A short question in a new chat will therefore usually use less AI usage than the same question in a long chat with several attachments. As a general rule, the more content, files, data, and processing steps involved, the more AI usage is required. ## **Monthly Usage Period** AI usage is measured within a monthly usage period. The period starts on the package start date and resets after one month. The product shows the percentage of included AI usage already used during the current period. Users can check their current usage level at any time. Admins receive early notifications when usage is high. This gives them time to decide whether a user should be moved to a Premium or Power profile, whether additional AI usage should be purchased or to just wait until the next monthly period. Once the AI usage has been fully used, further interactive AI actions are routed to the cost-efficient Instant model until the next monthly usage period begins. Users can continue working without interruption, keep access to AI features, and are not blocked. This does not create any automatic extra charges. AI usage is measured across all AI actions, including chat, the document editor, company knowledge, agents, apps, workflows, and chatbots. Workflows, API and chatbots are not automatically switched to another model, as this could affect the quality or result of a defined process. Once the AI usage has been used, further runs or API requests are paused until additional AI usage is purchased or approved, or until the next monthly usage period begins. ## **Usage in Day-to-Day Work** The included AI usage is designed for common tasks such as chat, knowledge search, summaries, meeting notes, documents, and the normal use of agents, workflows, and apps. Additional AI usage is generally required for long document reviews, complete presentations, large files, complex workflows, frequent automated runs, building and testing apps, API usage, or external chatbots. ## **User Profiles** Three user profiles are available for different usage levels: | **User profile** | **Suitable for** | **Included AI usage** | | :- | :- | :- | | Standard | Normal day-to-day usage | 1× | | Premium | Regular and more intensive usage | 2× | | Power | Very active users, AI Champions, and App Builders | 5× | The right profile depends on actual usage. Users with a consistently higher need can move from Standard to Premium or from Premium to Power. ## **Additional AI Usage** Additional AI usage can be purchased for more intensive use cases. This may be useful for larger apps, complex workflows, frequent automations, API connections, or external chatbots. Additional AI usage can be distributed flexibly across users and assigned where it is needed most. Extra charges only apply when the company explicitly purchases or approves additional AI usage. Unused AI usage is not carried over to the next monthly usage period. # User management Source: https://docs.nuwacom.ai/en/usermanagement Invite users and manage their access rights on the platform Effective user management ensures that all team members receive the appropriate access rights and can use the platform efficiently and securely. By assigning permissions precisely, you protect sensitive data and ensure compliance with data protection regulations. > [!NOTE] > > Only admins and superadmins have access to user management. *** ## **Invite and manage users** ![En Usermanagement](https://mintcdn.com/nuwacomgmbh/tgolVGcUQjuUn1QL/images/en_usermanagement.png?fit=max\&auto=format\&n=tgolVGcUQjuUn1QL\&q=85\&s=415a4a9e0f0e81905dbc18f6088e7686) In the workspace settings, open the "Users" section to manage your workspace members: * **Invite Users:** Add new team members by email invitation. * **Invitation Status:** See whether a user has accepted the invitation. If not, you can resend the email or remove the user. * **Assign Roles and Permissions:** Each user is assigned a specific role with corresponding permissions, determining their access to features and data in the nuwacom platform. Your rights depend on the role assigned to you – such as Admin, Editor, or Viewer. > [!NOTE] > > Note: Users must be invited to each workspace individually. *** ## **User Groups** ![Share Document En](https://mintcdn.com/nuwacomgmbh/cVOsCj_q2vK57TIf/images/Share_document_en.png?fit=max\&auto=format\&n=cVOsCj_q2vK57TIf\&q=85\&s=ad4c7f781169187a9d8d113ebb97b93c) With user groups, you can easily share items with entire teams. Instead of selecting individual users each time you share something, you bundle people into groups and use these groups when sharing chats, documents, projects, agents, and more. Only members of the selected group receive access to the respective item. **Your benefits at a glance** * Share content with a single user group instead of individual users. * Complete recurring sharing tasks in seconds. * Update the group once and automatically adjust access everywhere. **Example:**\ Share a knowledge folder by selecting the user group “Marketing Team” with the permission *Edit* in the sharing dialog. All members of this group will receive editing access to the folder. Other members of the workspace will not have access. **How it works** 1. Open the workspace settings. 2. In the left menu, navigate to *User Management → User Groups.* 3. Click *Create group* in the top right. 4. Enter a name for the user group. 5. Optional: Upload an icon or avatar for the group. 6. Click *Next.* 7. Add users via search with a click. 8. Click *Create group* to create the group. **Result:** The user group is now created and ready to be used for sharing chats, documents, projects, knowledge folders and more. ![Usermanagement2](https://mintcdn.com/nuwacomgmbh/tgolVGcUQjuUn1QL/images/usermanagement2.gif?s=91d291b6224435d62dcc6ad405acc065) **FAQ** > **Who can create and manage user groups?** > > User groups can be created and managed by admins and super admins. > **Can admins see all user groups?** > > Yes. Admins can see all user groups in the workspace, regardless of their own membership. > **When do I see a user group?** > > You see a user group if you are a member of that group or if you have admin rights. > **Am I automatically a member when I create a group?** > > Yes. The creator of a group is automatically added as a member. > **What permissions can I assign when sharing with a group?** > > When sharing an item, you can assign read or edit permissions, just as you would for individual users. These permissions apply to all members of the selected group. > > * **Read:** View content, but not edit it. > * **Edit:** View, edit, or delete content. > * **Remove access:** Revoke sharing at any time. > **What happens if group membership changes?** > > New members automatically gain access to all items shared with the group. Removed members lose access accordingly. *** ## **Roles and Permissions** The table gives you an overview of the roles and their respective access rights within the AI platform. Each role has specific permissions that regulate your access to certain features and data. > [!NOTE] > > The table shows the default permissions for each role. For Editors and Readers, these can be customized individually per workspace, and specific features can be activated or deactivated as needed. | **Feature** | **Action** | **Reader** | **Editor** | **Admin** | | :- | :- | :- | :- | :- | | **Chat** | Read | ✓ | ✓ | ✓ | | | Create / Update | ✓ | ✓ | ✓ | | | Delete | ✓ | ✓ | ✓ | | **Chat Documents** | Read | ✓ | ✓ | ✓ | | | Create / Update | ✓ | ✓ | ✓ | | | Delete | ✓ | ✓ | ✓ | | **Search** | Read | ✓ | ✓ | ✓ | | | Create Content | | ✓ | ✓ | | | Start Chat | ✓ | ✓ | ✓ | | **Project** | Read | ✓ | ✓ | ✓ | | | Create / Update | | ✓ | ✓ | | | Delete | | ✓ | ✓ | | **Document** | Read | ✓ | ✓ | ✓ | | | Create / Update | | ✓ | ✓ | | | Delete | | ✓ | ✓ | | **Meeting** | Read | ✓ | ✓ | ✓ | | | Create / Update | | ✓ | ✓ | | | Delete | | ✓ | ✓ | | **Agents** | Read | ✓ | ✓ | ✓ | | | Create / Update | | ✓ | ✓ | | | Delete | | ✓ | ✓ | | | Share | | ✓ | ✓ | | **Skills** | Read | ✓ | ✓ | ✓ | | | Create / Update | | ✓ | ✓ | | | Delete | | ✓ | ✓ | | | Share | | ✓ | ✓ | | **Prompts** | Read | ✓ | ✓ | ✓ | | | Create / Update | | ✓ | ✓ | | | Share | | ✓ | ✓ | | | Delete | | ✓ | ✓ | | **Tags** | Read | ✓ | ✓ | ✓ | | | Create / Update | | ✓ | ✓ | | | Delete | | ✓ | ✓ | | **Workflows** | Read | ✓ | ✓ | ✓ | | | Create / Update | | ✓ | ✓ | | | Delete | | ✓ | ✓ | | **Knowledge-Base Assets** | Read | ✓ | ✓ | ✓ | | | Create / Update | | ✓ | ✓ | | | Delete | | ✓ | ✓ | | **Knowledge Sync** | Read | ✓ | ✓ | ✓ | | | Sync to Agents or Projects | | ✓ | ✓ | | | Sync from Connectors | | | ✓ | | **User Groups** | Read | | | ✓ | | | Create / Update | | | ✓ | | | Delete | | | ✓ | | **Workspace Documents** | Read | ✓ | ✓ | ✓ | | | Create / Update | | ✓ | ✓ | | | Delete | | ✓ | ✓ | | **User Management** | Read | | | ✓ | | | Create / Update | | | ✓ | | | Delete | | | ✓ | | **Language Management** | Read | | | ✓ | | | Configure | | | ✓ | | **Model Settings** | Read | | | ✓ | | | Configure | | | ✓ | | **Analytics** | View | | | ✓ | | | Configure | | | ✓ | | **API Keys** | Read | | | ✓ | | | Create / Update | | | ✓ | | | Delete | | | ✓ | | **Apps** | Read | | | ✓ | | | Create / Update | | | ✓ | | | Delete | | | ✓ | | **Connectors (MCP/Credentials)** | Read | | | ✓ | | | Create / Update | | | ✓ | | | Delete | | | ✓ | | **Workspace** | Read | ✓ | ✓ | ✓ | | | Create / Update | | | ✓ | | | Delete | | | ✓ | *** ## **Manage Permissions** ![Rolematrix En](https://mintcdn.com/nuwacomgmbh/EZbccf0IzaHaONyB/images/rolematrix_en.png?fit=max\&auto=format\&n=EZbccf0IzaHaONyB\&q=85\&s=ad94d68b2d9ca425041259a947ffd109) As an Admin, you can use the permissions matrix to define which workspace features are available to Admins, Editors and Readers. Permissions can be activated or deactivated for each feature, for example for Agents, Workflows, Projects, Skills, or Prompts. This lets you control which roles can view, use, edit, or share specific areas. You can find the permissions under *Workspace Settings > User Management > Permissions.* Depending on the feature, different permissions are available, for example: | Permission | Meaning | | :- | :- | | Read | The feature or content can be viewed and used. | | Write | Content can be created or edited. | | Share | Content can be shared with other people. | ### Adjust permissions Activate or deactivate the desired permissions using the switches in the matrix. Then click ‘Save changes’ so that the adjustments are applied and take effect for the affected users. ### What happens when access is restricted? If a role does not have access to a feature, that feature is not displayed for the affected users or cannot be used. If a restricted feature is opened directly via a link, an access restriction is shown. # Videos Source: https://docs.nuwacom.ai/en/videos Create videos in seconds directly in nuwacom, without a camera or editing tools.\ Simply describe your video, optionally choose format and style, and seamlessly integrate the generated video into your workflows. With AI videos, you can: * Present content in a clear and visual way * Create marketing, recruiting, or product videos * Test and iterate ideas quickly * Save time and costs on video production ## **How to create a video** ![Promptgeneratevideo](https://mintcdn.com/nuwacomgmbh/tzlUXQ5D4ImeIx2T/images/promptgeneratevideo.png?fit=max\&auto=format\&n=tzlUXQ5D4ImeIx2T\&q=85\&s=75a9f15448b294cd303781021ca291e8) **1. Open chat** Open the chat and select “Generate video” to start creating your video. **2. Describe your video** Clearly describe what your video should look like, for example: *“Create a realistic video of two colleagues working together on a laptop. Modern office, natural interaction, smooth camera movement, professional look.”* **3. Set format and style (optional)** Optionally adjust format or style to further refine the result. **4. Send your request** Send your request and receive a finished video within seconds. ## **Chose format and style** You can control your video before generation: **Choose format** * Horizontal (16:9), ideal for presentations or websites * Vertical (9:16), suitable for social media **Define style** * Realistic * Cinematic * Motion graphics * Animated This allows you to tailor the video to your specific use case. ## **Integrate video into your workflow** After creation, you can directly continue working with your video and integrate it into your existing processes: **Download video**\ Download the finished video and use it for presentations, social media, or external communication. **Transfer content into a document**\ Move the video or its description into a document to further develop content, for example for scripts, campaigns, or internal materials. **Move to a project**\ Assign your video to a project to keep it organized and collaborate with your team. This ensures your video doesn’t remain isolated but becomes part of your existing workflows. ## **Share and collaborate** You can share the entire chat including the generated video with others. This allows colleagues to: * watch the video directly * understand the original prompt * create adjustments or new variations This is especially helpful for gathering feedback, collaboratively refining content, or sharing best practices within your team. ## **Models** The following models are available for video generation: * **Veo 2 (EU)** – EU-hosted, suitable for privacy-compliant usage * **Veo 3 (US)** – US-hosted > [!NOTE] > > By default, Veo 2 (EU) is used for video generation in your workspace. With the role 'Admin' you can change the model anytime in Workspace Settings > Models > Video Generation Model. This allows you to flexibly choose the model that best fits your use case. # Workspace settings Source: https://docs.nuwacom.ai/en/workspacesettings Organize and manage your workspace Workspace Settings provide a central location for configuring and managing a workspace. Admins can review the active plan and usage, define the workspace identity, manage available Workflows, access Analytics, configure Languages and Models, manage API Keys, administer Users and User Groups, connect external systems through Connectors, and apply workspace-wide AI Guardrails. Workspace Settings open through the settings gear icon in the top-left corner, next to the workspace name. The available settings are organized into the following areas: * **Plan & Usage:** Active plan, renewal date, included Seats by Seat-Type, consumption, and Additional AI Usage. * **General:** Workspace identity, including workspace name and logo. * **Workflows:** Availability of workflows within the workspace. * **Analytics:** Aggregated usage insights across Chat, Agents, Workflows, and the nuwacom API. * **Languages:** Available content languages and their display order. * **Models:** Default workspace model, image and video generation models, and the models enabled for users. * **API Keys:** Creation, scoping, and revocation of keys that authenticate requests to the nuwacom API. * **User Management:** Users, User Groups, Permissions, and SSO & User Sync. * **Connectors:** Apps and MCP Servers that link external systems to the workspace. * **AI Guardrails:** Voices, Guidelines, and Data protection that shape AI-generated content and govern data handling. Access to Workspace Settings is limited to admins. The available areas and settings depend on the workspace configuration and enabled features. *** # Organization ## Plan & Usage The Plan & Usage section is the central place for managing [AI usage](https://docs.nuwacom.ai/en/ai-usage) across the organization. Admins can monitor consumption, adjust usage limits, and assign Seats and Additional AI Usage. The section also displays the active plan, renewal date, and included Seats by Seat-Type. Admins can control usage at multiple levels: * Review and manage the plan, seats, and billing. * Monitor workspace and per-user AI usage. * Assign Seat-Types and Additional AI Usage to users. This gives admins flexible control over AI usage for teams and individual users in the organization. *** # Workspace ## **General** The General section contains the basic settings that define a workspace. These settings control the workspace identity and display information. The Change logo option uploads or replaces the workspace logo. A company, department, or project logo can be used. The Workspace Name field assigns a clear and descriptive name, such as "NUWA GmbH". *** ## Modules The module settings control which features are available in a workspace. ### **Workflows** Enables or disables the Workflows module for the workspace. When enabled, users in the workspace can access and use workflows. ### **Prompts** Enables or disables the Prompts module for the workspace. When enabled, users in the workspace can access and use prompts. > [!NOTE] > > Once a module is enabled, its features can be managed through permissions. Permissions control which features each role can use across the workspace. *** ## **Analytics** Analytics provides workspace administrators with aggregated insights into usage across Chat, Agents, Workflows, and the nuwacom API. Use it to monitor adoption, compare activity over time, and export chart data for reporting.\ \ The Analytics area includes the following tabs: * **Overview:** Provides central workspace metrics, including active users, messages, agents, workflows, and documents. * **Chat:** Shows Chat usage, including messages, model usage, and knowledge base searches. * **Agents:** Shows Agent usage, including Agent messages, active Agent users, messages per Agent, and model usage. * **Workflows:** Shows Workflow executions over time, including completed, failed, and cancelled executions. * **API:** Shows nuwacom API usage, including total API calls and calls over time. Select a date range and grouping to adjust the data shown in the charts. Hover over a chart to view exact values. Select *Export CSV* on a chart to download its aggregated data for further analysis or reporting. *** ## **Languages** The Languages section defines the languages available for workspace content. The order determines how languages are displayed to users. For example, set German as the primary language and English as the secondary language. New content is then created in German by default. The plus icon translates content into English. *** ## **Models** The Models section defines which [AI models](https://docs.nuwacom.ai/en/aimodels) are available in nuwacom. Admins control which models are enabled to users in the workspace. Auto Mode is recommended as the default setting. It automatically selects between a fast instant model and a more powerful model based on the complexity of each request. Team members receive a suitable response without selecting a model manually for every request. If required, a different available model can be selected manually in the chat. ### Default Workspace Model The default workspace model determines which model is preselected when a new chat starts. **How it works** * **Open**: Open Workspace Settings. * **Navigate**: Navigate to Workspace > Models. * **Expand**: Open the dropdown menu under Default Workspace Model. * **Select**: Select *Auto* or a specific model. The selected model is automatically preselected for new chats. > [!TIP] > > Select Auto to balance response quality, speed, and cost efficiency. ### Auto Mode Auto Mode selects the most appropriate model based on the complexity of the request. For simple tasks such as short answers, translations, or summaries, nuwacom uses a fast instant model. For more demanding tasks such as complex analyses, software development, or in-depth research, nuwacom uses a more powerful model. ### **Image Generation Model** Set the default model for image generation for your workspace. ### **Video Generation Model** Set the default model for video generation for your workspace. ### **Enabled Models** The Enabled Models section controls which language models are available to users in the workspace. * Filter models by provider, including OpenAI, Anthropic, Google, Meta, Mistral, and Cohere. * Select EU-hosted or US-hosted models to meet regional requirements and data protection preferences. * Enable or disable models using the activation toggles. ### Enabled Image Models The Enabled Image Models section controls which image generation models are available to users in the workspace. Workspace administrators select which models are enabled. Enable or disable image models using the activation toggles. Only enabled models are available for image generation in the workspace. ### Model Access Admins can limit an enabled language or image model to specific users, for example a model with extended retention or higher costs. * **Open**: Select the settings (gear) icon next to the model's toggle. * **Restrict access**: By default, *All Workspace Users* have access. Remove that entry to revoke access from everyone, then add back only the users who should have it by searching for their name or email. * **Save**: Select *Save access*. The model row then shows how many users have access. Users without access don't see the model in any model selection. With *All Workspace Users*, the model stays available to all users, including users who join the workspace later. If no one has access, nobody can use the model. * The default workspace model and the models used by Auto Mode are always available to everyone and can't be restricted. * Agents that use a restricted model run on the default workspace model for users without access. * Workflows run as their user. A workflow step that uses a restricted model fails when that user has no access. * Access is checked per user, including requests made with that user's API key. Requests from embedded agents are not limited by model access. [👉 View all available AI models](https://docs.nuwacom.ai/en/aimodels) *** ## API Keys API keys authenticate requests to the nuwacom API and serve as an alternative to password-based authentication. As an admin of the space, you can view and manage all API keys in the workspace, create new keys, and revoke existing keys. ### Create new API Key A new API key is created through the *Create new API Key* button. Each key is assigned a name and a set of permissions, and it can be revoked at any time. **How it works** * **Open**: Open Workspace Settings and select *API Keys*. * **Create**: Select *Create new API Key*. * **Name**: Enter a name for the key in the Name field. * **Set permissions**. Assign the required access level for each permission category, or use the *Set all* row to apply None, Read, or Write to all categories at once. * **Confirm**: Select *Create* to generate the key. ### **Permissions** Each API key is scoped through permissions that define the access level per area. Every category can be set to None, Read, or Write, and all categories default to None. *** # User Management User Management controls who has access to the workspace and what each user can do. It is available under Organization in Workspace Settings and lists all workspace users with their role, assigned Seat-Type, and account creation date. Access is limited to admins and superadmins. User Management is organized into four subsections: * **Users:** Invite users, assign roles and Seat-Types, and filter the user list. * **User Groups:** Bundle users into groups for shared access and simpler management. * **Permissions:** Define which features and areas each role or group can access. * **SSO & User Sync:** Configure Single Sign-On and automated user synchronization with an identity provider. *** ## Users The Users section manages all workspace members, including invitations, invitation status, and the assignment of roles and Seat-Types. * **[Invite users](https://docs.nuwacom.ai/en/usermanagement#invite-and-manage-users):** Add members through an email invitation. * **Invitation status:** Check whether an invitation was accepted, and resend or remove pending invitations. * **Assign roles and Seat-Types:** Each user receives a role that determines access to features and data, for example Admin, Editor, or Viewer. > [!NOTE] > > Users are invited to each workspace individually. *** ## User Groups [User Groups](https://docs.nuwacom.ai/en/usermanagement#user-groups) bundle multiple users so items are shared with entire teams at once. When sharing a chat, document, project, agent, or knowledge folder, a group is selected instead of individual users, and only its members receive access. Updating a group once adjusts access everywhere the group is used. *** ## Permissions Permissions define which features and areas each role or group can access. Admins use the permissions matrix to control access to features such as Agents, Workflows, Projects, Skills, and Prompts, and to determine which roles can view, use, edit, or share each area. The matrix shows the default permissions per role. For Editors and Readers, these defaults can be customized per workspace. Depending on the feature, different permission levels are available: | Permission | Meaning | | :- | :- | | Read | The feature or content can be viewed and used. | | Write | Content can be created or edited. | | Share | Content can be shared with other people. | > [!NOTE] > > Admins always retain full access and cannot be restricted. *** ### SSO & User Sync [SSO & User Sync](https://docs.nuwacom.ai/en/sso-and-usersync) configures Single Sign-On and automated user synchronization with an identity provider. Members authenticate through the connected identity provider, and accounts stay aligned automatically as the provider is updated. *** # Connectors nuwacom lets you connect AI with the tools your team already uses. Native Connectors make information and actions from your existing tool stack available across nuwacom modules such as Agents, Workflows, and Apps. In the settings, admins can enable or disable individual Connectors. By default, all Connectors are enabled. Admins can manage each Connector, connect accounts, and define whether it is private, available to selected team members, or shared with the entire workspace. *** ## **Apps** pps are preconfigured integrations for common tools such as email, calendars, and collaboration platforms. They can be set up quickly and allow agents to analyze content, retrieve information, or perform actions. App availability is managed centrally, and each connector can be enabled or disabled through the activation toggles. **How it works** * **Open**: Open Workspace Settings. * **Select**: Select *Apps*. * **Toggle:** Enable or disable apps using the activation toggles. * **Authenticate**: Complete the authentication process. App access can remain private or be shared with selected team members or the entire workspace. Sharing a connected account grants access to the data available through that connector. *** ## **MCP Servers** MCP servers provide deeper and more flexible integrations than Apps. The MCP standard connects almost any system and makes complex actions available within agents and workflows. MCP servers are intended primarily for admins and advanced teams implementing custom automations and system integrations. MCP servers are primarily intended for administrators and advanced teams that want to implement custom automations and system integrations. **How it works** * **Open**. Open Workspace Settings, select *Connectors*, and select *Add Server*. * **Verify**. Enter the server URL and, optionally, a display name, then select *Check*. nuwacom inspects the server and detects the transport type (StreamableHTTP or SSE), whether authentication is required, and which method is used. * **Authenticate**. Name the connection and enter the credentials for the detected authentication method (see Authentication Methods). * **Test**: Select *Test Connection*. nuwacom connects to the server and detects the available tools. If the test fails, the error returned by the server is displayed so the settings can be adjusted. * **Select tools**: Select the tools the agents are allowed to use, then confirm to create the connection. *** # AI Guardrails AI Guardrails define communication styles, brand guidelines, and quality standards across the workspace. The configured rules are automatically applied to AI-generated responses. *** ## **Voices** Voices define the tone and writing style used across the workspace. Admins can add, manage, and analyze voices for different purposes, such as CEO speeches, social media, or internal communication. Once created, a voice is available to all AI-powered writing tools and helps keep messaging consistent and brand-compliant. The Voices section is available under AI Guardrails and includes ready-to-use Default Voices such as Confident, Direct, Empathetic, Friendly, and Trustworthy, alongside any custom voices. New voices are created through Add a New Voice, using one of two methods: * **Identify Your Voice (Recommended):** AI analyzes a document, a URL, or entered content and creates a tone summary that can be adjusted. * **Enter Your Voice Manually:** The tone is described directly, using characteristics such as formality, emotional intensity, subjectivity, and consistency. **How it works** * **Open**: Open Workspace Settings and navigate to AI Guardrails > Voices. * **Add**: Select *Add a New Voice*. * **Choose a method**: Select *Identify Your Voice* or *Enter Your Voice Manually*. * **Provide input**: Upload the source, enter the URL, or describe the tone. * **Save**: Confirm to add the voice, then adjust the result if required. *** ## Guidelines Guidelines define the general knowledge that shapes every AI-generated response in the workspace. Admins enter this knowledge once, and it is automatically applied to answers, social media posts, blog articles, and other content, so output follows the defined communication style, wording, and brand rules. The Guidelines section is available under AI Guardrails and provides a single text area that supports Markdown formatting, allowing admins to structure content with headings, lists, and emphasis. Typical guideline content includes: * Company overview * Brand personality and rules * Writing style and tone * Wording and terms to avoid * Facts and reference information **How it works** * **Open**: Open Workspace Settings and navigate to AI Guardrails > Guidelines. * **Enter**: Enter the guidelines in the text area, using Markdown to structure the content. * **Save**: Select *Save* to apply the entries across the workspace. > **[Download Guidelines-Template](https://nuwacom.com/wp-content/uploads/2025/09/AI-Guardrails-%E2%80%93-Guidelines.pdf)** > > This document is a sample template for Guidelines. It shows how you can define your own rules for language, style, and wording to ensure that AI-generated content is consistent and brand-compliant. Use it as a template and add your company-specific guidelines. 👉 [Read more](https://docs.nuwacom.ai/en/policy-hierarchy) to understand how policies, knowledge sources, and system instructions are prioritized when you submit a prompt. *** ## **Privacy Mode** Privacy Mode protects personal data before it reaches an AI model. When enabled, nuwacom automatically detects and replaces sensitive information — such as names, email addresses, and IBANs — with neutral placeholders. The AI processes only those placeholders. Once a response is generated, the original values are restored before users see the result. Privacy Mode is available under AI Guardrails in the workspace settings. *Enforce Privacy Mode* for all users controls whether the protection applies workspace-wide. When active, Privacy Mode is always on and cannot be turned off in chat by users. When *Enforce Privacy Mode for all users* is inactive, each user can manage Privacy Mode for their own messages using the *Turn on for my messages only* toggle in the chat composer. Turning it on protects that user’s messages only; other workspace members are unaffected. # Chat Source: https://docs.nuwacom.ai/product/chat Interactive chat for creating content and answering questions The AI chat from nuwacom helps you use internal data, connected apps, or web search to research, develop ideas, and create or edit content, all in one interface. Access AI apps, agents, or skills at any time to support your work, and share your chats and content with your team to continue developing them together. ## Features at a glance Discover how to integrate the AI-powered chat into your daily work – for research, documents, emails, and more. > **Enterprise Search** > > Search across all relevant data sources in one place: the web for up-to-date information, your internal knowledge for answers from documents, and connected apps like OneDrive, SharePoint, or Outlook. > **Tasks** > > Create slides, documents, notes, and emails, and complete your key tasks in one place. You can also research on the web, generate images and videos, analyze data, build workflows, and create your own agents. > **Type @ to reach all your AI** > > Simply type `@` while writing to open a menu with your AI apps, agents, skills, prompts, voices, and workflows, and use them directly in chat. > **Workflows** > > Automate recurring tasks by combining multiple steps into a workflow. You control the logic in the chat while your workflow is built in parallel in the editor. > **Orchestrator** > > Handle complex tasks with the Orchestrator. It has multiple skills and can orchestrate agents and tools. It finds relevant data and executes tasks step by step – across documents, slides, visuals, and workflows. > **LLM-agnostic Chat** > > Interact with the latest and most powerful LLMs in one interface. In nuwacom, you can choose models like GPT-5.5, Claude Opus 4.8, or Gemini 3.1 Pro Preview – for text, data analysis, and multimodal tasks. > **Attach files and images** > > Add relevant files, templates, images, or knowledge to give the AI more context. This allows it to generate content directly based on your data – in the right format, layout, and brand style. > **Collaboration** > > Easily share chats with your team and build on existing results together – across your entire workspace or with specific individuals. *** ## **Input field** ![Chaten](https://mintcdn.com/nuwacomgmbh/tzlUXQ5D4ImeIx2T/images/chaten.png?fit=max\&auto=format\&n=tzlUXQ5D4ImeIx2T\&q=85\&s=cf3d171557141b0fdbdb09bd5f12d795) **How to interact with the AI: type or speak your request** The input field is your central interface for interacting with the AI. Enter your prompt by typing directly into the field, or click the microphone icon to use voice input. Both methods convert your input into actionable requests for the AI. **Explore the possibilities** * **Ask questions:** Get answers based on your internal knowledge base, the web, or connected apps. To activate a source, click the + icon in the input field and enable the desired source (internal knowledge, web, or apps).\ **Example:** *“What are the key arguments for our product compared to competitors?”* * **Access up-to-date information:** Retrieve current information from the internet by asking a question. Web search is enabled by default and used automatically.\ **Example:** *“What are the three key trends in B2B content marketing in the software industry right now?”* * **Real results:** Start your work directly in the chat and create slides, documents, notes, emails, and more in one place. The AI provides results as downloadable files such as presentations (.pptx), Excel sheets (.xlsx), or text files (.txt). * **Automate tasks:** Use the orchestrator or workflows to efficiently handle recurring or complex tasks.\ **Example:** *“Create a landing page structure for our new campaign including H1, CTA, and bullet points.”* **💡Tip:** Try different phrasings, combine sources, or start tasks directly in the chat to get the best results from your prompt. *** ## **Actions in chat** ![Aimenuen](https://mintcdn.com/nuwacomgmbh/FBUzIFKQ_xmikLDJ/images/aimenuen.png?fit=max\&auto=format\&n=FBUzIFKQ_xmikLDJ\&q=85\&s=0d844f2a6a9c9681a40e1f4a966ba876) Use actions in chat to call agents, skills, and saved prompts directly from your conversation or to add additional content. **Use the @ symbol to start relevant actions such as apps, agents, workflows, skills, language styles, or prompts**. Use the + button to add files and images, select content from the knowledge base, or search connected sources and integrations. This lets you work in chat with exactly the functions and information you need for your task, without switching between different areas. ### **Call an action with the @ symbol** 1. Open a chat. 2. Type @ in the input field. 3. Select the action you want to use. For example, you can call the following actions: * **Agents**: Start specialized assistants for specific tasks. * **Skills**: Use predefined capabilities, for example to create documents, slides, or other results. * **Prompts**: Use saved prompts from your prompt library. ### **Add files and sources with the + button** The + button complements actions in chat. You can use it to attach files or images and select sources for your request. With the + button, you can, for example: * Upload files * Attach images * Select content from the knowledge base * Search connected sources and integrations ### Use multiple actions You can type @ again at any time to add more actions. This is helpful if you realize while working that you also need a specific agent, skill, or prompt. Example: You first start an agent for research and then use a saved prompt to turn the results into a LinkedIn post. *** ## **Tasks** ![Chattasksen](https://mintcdn.com/nuwacomgmbh/tzlUXQ5D4ImeIx2T/images/Chattasksen.png?fit=max\&auto=format\&n=tzlUXQ5D4ImeIx2T\&q=85\&s=eec60464b42d3e9bf56a9b207da2cf2a) Start typical work processes directly in the chat. Simply choose the right task to structure your request and get real results, for example: * **Use Orchestrator** – Start complex tasks that are automatically planned, coordinated, and handled by specialized agents. * **Create slides** – Generate presentation slides for meetings or pitches. * **Create document** – Write structured content such as concepts or reports. * **Build workflows** – Automate recurring processes by connecting multiple steps. * **Write notes** – Capture ideas or meeting notes. * **Research** – Conduct structured research on markets, competitors, or trends. * **Search company knowledge** – Explore company knowledge and find relevant information from connected sources. * **Generate images** – Create visual content from a description. * **Generate videos** – Create videos based on text descriptions. * **Analyze data** – Evaluate or summarize datasets. * **Create agent** – Build your own AI agents to support recurring tasks. * **Draft email** – Write professional emails. You can select a task directly below the input field and pin frequently used tasks. *** ## **Model selection** ![Group 14710](https://mintcdn.com/nuwacomgmbh/osV_aASm-7aDJPwy/images/Group-14710-1.png?fit=max\&auto=format\&n=osV_aASm-7aDJPwy\&q=85\&s=6a7a7be3a5921635d8b3e6d9bf64295e) Auto Mode is enabled by default in the chat. nuwacom automatically selects the model that best fits each request, providing fast responses for simple tasks and more advanced capabilities for complex ones. If needed, you can switch to a different model at any time using the dropdown menu. nuwacom provides access to a wide range of modern AI models that are flexible, securely hosted, and optimized for real-world use cases. * **Auto Mode:** Delivers the right balance of speed, quality, and resource efficiency. Simple requests are handled quickly, while more complex tasks are routed to a more powerful model. * **Manual Model Selection:** Alternatively, you can choose a specific model from the model dropdown. This selection overrides Auto Mode for the current request. * **Latest AI Models from Leading Providers:** Choose from powerful models such as GPT-5.5, Claude Opus 4.8, or Gemini 3.1 Pro Preview, ideal for text generation, data analysis, and multimodal tasks. * **Flexible Hosting Options:** Choose between EU-hosted models (GDPR-compliant and privacy-focused) and US-hosted models. * **Available Even During Provider Outages:** If an LLM provider becomes temporarily unavailable, your chat continues through an automatic multi-provider fallback chain. All fallback models remain EU-hosted. * **The Right Model for Every Task:** Whether you need fast responses for routine work, advanced reasoning for complex topics, or support for visual content, there is a suitable model for every scenario. 👉 [Learn more about AI models](https://docs.nuwacom.ai/en/aimodels) *** ## **Attach files and images** ![Chatuploaden](https://mintcdn.com/nuwacomgmbh/tzlUXQ5D4ImeIx2T/images/chatuploaden.png?fit=max\&auto=format\&n=tzlUXQ5D4ImeIx2T\&q=85\&s=2f03e94b318609470da12660979c30fd) Provide the AI with targeted context by adding relevant documents via plus icon: * **Drag and drop files into the chat** (PDF, Word, Excel)\ *Example:* Drag your product brochure into the chat – the AI can use it to create a clean press release. * **Upload presentations, guides, or FAQs**\ *Example:* Upload your latest campaign analysis PPT so the AI can create a management report. * **Use your own templates for slides and documents**\ *Example:* Upload your presentation or document template so the AI creates new content in the correct layout and brand style. * **Upload images for content or visual analysis**\ *Example:* Add a diagram – the AI can describe, summarize, or integrate it into supporting text. * **Attach files**\ *Example:* Use your style guide from the knowledge base so the AI writes in your exact brand voice. * **Add content from nuwacom**\ *Example:* Use content you created in nuwacom, such as documents, emails, slides, or meeting notes. * **Connect data from OneDrive and Google Drive**\ Access content from your cloud storage – such as presentations, briefings, or reports – and automatically provide the AI with the necessary information. > **Supported file formats** > > **Plain text & markup** > > * `application/json` > * `text/plain` > * `text/markdown` > * `text/html` > * `application/xml` > * `text/xml` > * `text/x-rst` > * `application/rtf` > * `text/rtf` > > **Emails & messages** > > * `application/vnd.ms-outlook` > * `message/rfc822` > > **Images** > > * `image/jpeg` > * `image/png` > * `image/webp` > * `image/gif` (non-animated) > > **Documents (Office & standard)** > > * `application/msword` > * `application/vnd.openxmlformats-officedocument.wordprocessingml.document` > * `application/vnd.oasis.opendocument.text` > * `application/epub+zip` > > **Tables & spreadsheets** > > * `text/csv` > * `application/vnd.ms-excel` > * `application/vnd.openxmlformats-officedocument.spreadsheetml.sheet` > > **Presentations** > > * `application/vnd.ms-powerpoint` > * `application/vnd.openxmlformats-officedocument.presentationml.presentation` > > **PDF** > > * `application/pdf` > > **Audio & Video** > > * `audio/mpeg` > * `audio/wav` > * `audio/ogg` > * `audio/flac` > * `audio/aac` > * `audio/aiff` > * `audio/mp4` > * `video/mp4` > * `video/webm` > > **Google Drive** > > * `application/vnd.google-apps.document` > * `application/vnd.google-apps.spreadsheet` > * `application/vnd.google-apps.presentation` > > **nuwacom documents** > > * nuwacom documents > * nuwacom meeting notes *** ## **Enterprise Search** ![Chatselectsourceen](https://mintcdn.com/nuwacomgmbh/tzlUXQ5D4ImeIx2T/images/chatselectsourceen.png?fit=max\&auto=format\&n=tzlUXQ5D4ImeIx2T\&q=85\&s=bd3098a98f80b1699abc02e56d96129b) Define where the AI should search for information to answer your request. You can choose via plus-icon whether the AI uses the web, your internal knowledge, or connected applications. Web search is enabled by default. **Web (public internet)**\ Use web search for up-to-date external information such as trends, market analysis, or competitors.\ Example: “Which TikTok hashtags are trending in the beauty industry?”\ → The AI scans current rankings, articles, and trend reports. **Internal knowledge (your company data)**\ Use internal knowledge for content that exists only within your company.\ Example: “Which headline do we use in the summer campaign?”\ → The AI accesses your style guide and previous campaign documents. **Connected apps (e.g. OneDrive, SharePoint, Outlook)**\ Use connected apps when information is stored in your daily tools.\ Example: “Summarize the key points from the project briefing.”\ → The AI searches relevant documents, emails, or files in your connected systems. **Important:**\ You can activate multiple sources at the same time. *** ## Work in a Folder A chat can carry its own working folder. Connecting a Knowledge Folder to the chat links the conversation to a specific location in nuwacom, where existing files are available as context and new files are stored. This gives ongoing work a clear home: source material, chat findings, and generated files all live in one place, ready for reuse across the team. ### How it works * **No folder selected:** The chat is not connected to any folder. Existing folder files do not inform the conversation, and no new files are stored there. * **Folder selected:** Files in the folder are available as context, and new files are stored there. The selection stays active across all prompts in the session. * **Switching folders:** A different folder can be selected at any time. It replaces the previous one and becomes the active context and storage location for subsequent prompts. * **Remove a folder:** Hovering over the selected folder displays the X. Selecting the X removes the folder, and the chip returns to “Folder.” ### Using the folder as context With a folder active, the files inside it become available as context for the conversation. nuwacom draws on that material to answer prompts, so questions can reference documents in the folder without re-uploading them. *Example:* A customer success manager selects the "Customer Acme" folder containing the contract, meeting notes, project documentation, and previous support summaries, then asks nuwacom to summarize the current customer status and highlight open topics. **How it works** 1. Open a new chat. 2. Select a folder, such as "Customer Acme". 3. Ask a question about the files in the folder, for example: "Summarize the current customer status and list the open topics and next steps." 4. Review the response. nuwacom uses the selected folder’s files as context. > [!NOTE] > > **Note:** Only folders with edit rights are available for selection. ### Using the folder to save files Everything produced or uploaded in the chat is saved to the selected folder in Knowledge, including: * Documents and slide decks created in the chat. * Other generated files, such as spreadsheets and PDFs. * Files uploaded into the chat. *Example:* After reviewing the customer information, the manager asks nuwacom to create a meeting briefing for the next customer call. The generated document is saved directly in the "Customer Acme" folder alongside the source material. **How it works** 1. Stay in the chat connected to the "Customer Acme" folder. 2. Ask nuwacom to create a file, for example: "Create a meeting briefing document with the current status, open topics, and next steps." 3. Review the generated file. 4. The file is saved automatically in the "Customer Acme" folder alongside the source documents. 5. You or a colleague can open, review, and reuse it later. *** ## **Run agents** Use agents to work through more complex tasks directly in chat in a structured way. An agent guides you through the process step by step, asks targeted questions, and takes stored knowledge, instructions, or documents into account when needed. Call the right agent via the @ menu. To do this, type @, select agents, and search for the desired agent. You can then start it directly in your current conversation. **Example:** Start a social media agent that is linked to your brand guide. The agent asks for target audience, platform, and topic, and creates a finished post in the right brand style, including caption and hashtags. *** ## **Run skills** With skills, you control how nuwacom responds to your request. For example, they define a specific structure, tone, or fixed output format, ensuring consistent results. Call the desired skill via the @ menu by typing @, selecting skills, and searching for the right skill. Then formulate your task directly in chat. **Example:** Select the “Competitive analysis” skill and write: “Analyze how our competitors are positioned in the market.” nuwacom then creates the response in the defined format of the skill. *** ## **Run prompts** Use prompts to complete recurring tasks faster and more precisely. A prompt gives your request a clear direction and helps you get better results right away, without having to rewrite the right wording every time. Call the right prompt directly in chat via the @ menu. To do this, type @, select prompts, and search for the desired prompt from the library. You can then use it directly in your current conversation. **Example:** Select the prompt “Check grammar & spelling”, paste your text, and have it automatically checked for spelling and grammar. *** ## **Apply voices** Use language styles to create content faster in the right tone. A language style ensures that texts sound consistent, match the brand, and do not require you to explain every time how a text should be phrased. Call the right language style directly in chat via the @ menu. Type @, search for the language style, and mention it in your message. nuwacom then applies the style to your request. **Example:** Mention the language style “Brand communication” and write: “Write a LinkedIn post about our new feature.” nuwacom then creates the text in the selected language style. *** ## **Run workflows** Use workflows to run recurring processes faster and more reliably. A workflow bundles multiple steps, agents, or tools into a fixed sequence that you do not have to restart or coordinate manually every time. Call the right workflow directly in chat via the @ menu. Type @, search for the workflow, and mention it in your message. The workflow starts automatically and executes the defined process. **Example:** Mention the workflow “Calendar briefing” to have your day automatically structured before your first appointment. *** ## **Source references** Maintain full transparency over the quality of your AI responses – all sources used are traceable and directly accessible. The output shows which sources the AI used. This allows you to choose data sources intentionally and get more precise answers. * **Full transparency** – all used internal documents and web links are displayed directly below the response. * **Instant fact-checking** – click a source to open the original file or webpage and verify information quickly. * **Flexible source control** – use the source menu to view and adjust selected data sources for higher relevance and accuracy. * **Higher credibility** – traceable sources strengthen presentations, analyses, and whitepapers. > [!NOTE] > > A corresponding citation is shown whenever information from a source is reflected directly or indirectly in the AI-generated response. Sources that contribute to the answer without being directly or indirectly cited may still influence the response and are therefore also listed under “Sources”. *** ## **Continue working with AI outputs** Generated outputs can be edited, refined, and collaboratively used within your team in just a few clicks. Below, you’ll find a selection of options – each with a practical example: * **Copy and reuse output**\ Copy generated text and use it directly in other tools – such as emails, presentations, or internal documents.\ Example: Copy a generated campaign description and insert it into a sales presentation. * **Share output**\ Easily share content with your team for review or approval. Example: Turn a generated campaign idea into a full blog article and align it with your team. * **Real results**\ Turn AI responses directly into emails, documents, or presentations. The AI automatically structures the content so you can use, edit, or send it immediately.\ *Example:* Turn a meeting summary into an email with key outcomes and next steps. * **Generate files**\ Create ready-to-use files from AI outputs in formats such as presentations (.pptx), Excel sheets (.xlsx), or text files (.txt). The file is generated in the chat and available for download.\ Example: Create a presentation for a strategy meeting or an Excel file for competitor analysis. *** # Skills in Agents Source: https://docs.nuwacom.ai/skills-in-agents A Skill can be attached directly to an Agent to define how the Agent responds. Once a Skill is added, it activates automatically for everyone using that Agent, without individual installation or selection. This produces fully configured Agents with tone, structure, and rules already built in. Every output follows the same standard from the first interaction onward. **Example** An Agent with a *Meeting Notes Skill* structures every output into Key Decisions, Open Questions, and Next Steps. No setup is required on the user's end. ### Skills vs. Agents | | Skill | Agent | | :- | :- | :- | | Defines | How nuwacom responds | What nuwacom does | | Focus | Format, tone, structure | End-to-end task execution | | Mode | Passive: activates automatically or on demand | Active: runs autonomously | | Think of it as | The rulebook | The executor | Applying a Skill to an Agent combines autonomous task execution with guaranteed output quality. Any behavior that should stay consistent, regardless of who initiates it, belongs in a Skill. # Create an agent Source: https://docs.nuwacom.ai/api-reference/agents/create-an-agent /openapi.yaml post /api/v1/spaces/{spaceId}/agents Create a new agent in a space. The agent starts as an unpublished draft (`published: false`); publish it in the app to share it with the space. # Delete an agent Source: https://docs.nuwacom.ai/api-reference/agents/delete-an-agent /openapi.yaml delete /api/v1/spaces/{spaceId}/agents/{agentId} Permanently delete an agent, including its draft and published settings, actions and skills. # Get an agent Source: https://docs.nuwacom.ai/api-reference/agents/get-an-agent /openapi.yaml get /api/v1/spaces/{spaceId}/agents/{agentId} Return a single agent. Built-in nuwacom system agents are not available through this endpoint (404). The returned fields reflect the draft state (the version this API edits); `published` tells whether a published version exists. # List agents Source: https://docs.nuwacom.ai/api-reference/agents/list-agents /openapi.yaml get /api/v1/spaces/{spaceId}/agents List the agents of a space, ordered by most recently updated first. Built-in nuwacom system agents are excluded — only agents created in the space (via this API or the app) are returned. Results are paginated via the `limit` and `offset` query parameters; the response wraps the agents in `data` alongside `total`, `limit`, `offset`, and `hasMore`. The returned fields reflect the draft state (the version this API edits); `published` tells whether a published version exists. Use the agent ID with the [Completion API](/api-reference/completion-api/openai-chat-completion) (`agentId` option) to chat with an agent. # Publish an agent Source: https://docs.nuwacom.ai/api-reference/agents/publish-an-agent /openapi.yaml post /api/v1/spaces/{spaceId}/agents/{agentId}/publish Make the draft the published version of the agent. Copies the draft settings (including actions and skills) over the published version, so the Completion API uses them by default. If the draft has `embedEnabled: true`, publishing also makes the external embed widget live; this requires a non-empty `embedOptions.domainWhitelist`. Who can see and use the agent in-app is controlled separately via sharing in the nuwacom app. Idempotent: publishing an already published agent without draft changes is a no-op. # Unpublish an agent Source: https://docs.nuwacom.ai/api-reference/agents/unpublish-an-agent /openapi.yaml post /api/v1/spaces/{spaceId}/agents/{agentId}/unpublish Remove the published version of the agent; it reverts to a draft. The draft keeps all settings and can be re-published at any time. Published copies of actions and skills are removed along with the published version. Idempotent: unpublishing an unpublished agent is a no-op. # Update an agent Source: https://docs.nuwacom.ai/api-reference/agents/update-an-agent /openapi.yaml patch /api/v1/spaces/{spaceId}/agents/{agentId} Update an agent. Changes are written to the draft; the published version stays untouched until the agent is re-published in the app (`hasUnpublishedChanges` flags the drift). Omitted fields keep their current value. # List all AI models Source: https://docs.nuwacom.ai/api-reference/ai/list-all-ai-models /openapi.yaml get /api/ai/models List all supported AI models # List space AI models Source: https://docs.nuwacom.ai/api-reference/ai/list-space-ai-models /openapi.yaml get /api/ai/models/{spaceId} List all supported AI models for a specific space # Add an asset via URL Source: https://docs.nuwacom.ai/api-reference/assets/add-an-asset-via-url /openapi.yaml post /api/kb/assets Add an asset to the system by providing a URL to the asset # Delete an asset Source: https://docs.nuwacom.ai/api-reference/assets/delete-an-asset /openapi.yaml delete /api/kb/assets/{spaceId}/{assetId} Permanently delete an asset from the given space. This removes the stored file and any derived embeddings for that asset. # Get an asset Source: https://docs.nuwacom.ai/api-reference/assets/get-an-asset /openapi.yaml get /api/kb/assets/{spaceId}/{assetId} Return the complete asset record (name, MIME type, connector info, metadata, processing state, etc.) for the provided asset as long as the caller has read access to the space. # Update an asset Source: https://docs.nuwacom.ai/api-reference/assets/update-an-asset /openapi.yaml patch /api/kb/assets/{spaceId}/{assetId} Update mutable asset metadata (name, connector assignment, metadata JSON, file path info, etc.) without re-uploading the file itself. # Upload an asset Source: https://docs.nuwacom.ai/api-reference/assets/upload-an-asset /openapi.yaml post /api/kb/assets/upload Upload an asset to the system # OpenAI Chat Completion Source: https://docs.nuwacom.ai/api-reference/completion-api/openai-chat-completion /openapi.yaml post /api/v1/openai/chat/completions Create a new AI completion. You can find the supported models using [List all AI models](/api-reference/ai/list-all-ai-models). # Create a skill Source: https://docs.nuwacom.ai/api-reference/skills/create-a-skill /openapi.yaml post /api/v1/spaces/{spaceId}/skills Create a new skill in a space. The `instructions` become the body of the skill's SKILL.md file (its frontmatter is generated from `name` and `description`). The skill is active immediately; attach it to agents via the Agents API or in the nuwacom app. # Create or replace a skill file Source: https://docs.nuwacom.ai/api-reference/skills/create-or-replace-a-skill-file /openapi.yaml put /api/v1/spaces/{spaceId}/skills/{skillId}/files/{path} Create a new text file in the skill's folder or fully replace an existing one. Parent folders are created implicitly by the path. Writing to `SKILL.md` replaces the whole file including its frontmatter; when the new frontmatter contains a valid `description`, the skill's metadata is synced to it. Prefer [Update a skill](/api-reference/skills/update-a-skill) with `instructions` for SKILL.md changes — it keeps the frontmatter consistent automatically. Only text files can be written; binary assets are not supported via the API. # Delete a skill Source: https://docs.nuwacom.ai/api-reference/skills/delete-a-skill /openapi.yaml delete /api/v1/spaces/{spaceId}/skills/{skillId} Permanently delete a skill, including its SKILL.md and any other files. Agents that had the skill attached simply lose it. Built-in nuwacom skills cannot be deleted. # Delete a skill file Source: https://docs.nuwacom.ai/api-reference/skills/delete-a-skill-file /openapi.yaml delete /api/v1/spaces/{spaceId}/skills/{skillId}/files/{path} Delete a file from the skill's folder. `SKILL.md` cannot be deleted — every skill needs one; change it via [Update a skill](/api-reference/skills/update-a-skill) instead. # Disable a skill Source: https://docs.nuwacom.ai/api-reference/skills/disable-a-skill /openapi.yaml post /api/v1/spaces/{spaceId}/skills/{skillId}/disable Disable a skill for the whole space: the AI stops loading it for everyone (users can still re-enable it for themselves in the nuwacom app). The skill and its files stay untouched — use this instead of deleting when you want to switch a skill off temporarily. Re-activate it via [Enable a skill](/api-reference/skills/enable-a-skill). # Enable a skill Source: https://docs.nuwacom.ai/api-reference/skills/enable-a-skill /openapi.yaml post /api/v1/spaces/{spaceId}/skills/{skillId}/enable Enable a skill for the whole space, so the AI loads it whenever its description matches the task. Skills are enabled by default when created; use this to revert a [Disable](/api-reference/skills/disable-a-skill). A space can have at most 40 active skills — enabling beyond that returns 409. Individual users can still override the space-wide state for themselves in the nuwacom app. # Get a skill Source: https://docs.nuwacom.ai/api-reference/skills/get-a-skill /openapi.yaml get /api/v1/spaces/{spaceId}/skills/{skillId} Return a single skill, including its `instructions` (the body of the skill's SKILL.md file). # Get a skill file Source: https://docs.nuwacom.ai/api-reference/skills/get-a-skill-file /openapi.yaml get /api/v1/spaces/{spaceId}/skills/{skillId}/files/{path} Return the text content of a file in the skill's folder. Only text files can be read through this endpoint; binary assets (images, fonts) are not available via the API. # List skill files Source: https://docs.nuwacom.ai/api-reference/skills/list-skill-files /openapi.yaml get /api/v1/spaces/{spaceId}/skills/{skillId}/files List all files in the skill's folder, including its SKILL.md. Use the returned `path` values with [Get a skill file](/api-reference/skills/get-a-skill-file). # List skills Source: https://docs.nuwacom.ai/api-reference/skills/list-skills /openapi.yaml get /api/v1/spaces/{spaceId}/skills List the skills of a space, ordered by most recently updated first. Results are paginated via the `limit` and `offset` query parameters; the response wraps the skills in `data` alongside `total`, `limit`, `offset`, and `hasMore`. Only user-created skills are returned; built-in nuwacom skills are not exposed through this API. Use [Get a skill](/api-reference/skills/get-a-skill) to also read the skill's `instructions`. # Update a skill Source: https://docs.nuwacom.ai/api-reference/skills/update-a-skill /openapi.yaml patch /api/v1/spaces/{spaceId}/skills/{skillId} Update a skill. Omitted fields keep their current value. When `instructions` is provided it fully replaces the body of the skill's SKILL.md file; name/description changes are also synced into the SKILL.md frontmatter. Built-in nuwacom skills cannot be updated.