Outline MCP Server MCP Server
io.github.Vortiago/mcp-outline
Connect AI assistants to Outline for document search, reading, and management
What is the Outline MCP Server MCP server?
The Outline MCP Server is a Model Context Protocol server that enables AI assistants to interact with Outline document management systems. It provides tools for searching, reading, creating, editing, and managing documents, collections, and comments within Outline, with support for both cloud-hosted and self-hosted instances.
This server bridges AI assistants like Claude and Cursor to your Outline knowledge base, enabling natural language document search, content reading, document creation and editing, collection management, and collaborative features like comments. It includes automatic rate limiting, caching, and flexible access control modes (read-only, disable-delete) suitable for multi-user deployments.
How to install Outline MCP Server
Copy-paste configuration for popular MCP clients.
OUTLINE_API_KEYsecretOutline API key (Settings > API). Can also be set in ~/.config/mcp-outline/.env
OUTLINE_API_URLOutline API URL for self-hosted instances
Tools & capabilities
Tools this server exposes to the agent.
search_documents— Search documents by keywords with pagination and status filteringlist_recently_updated_documents— List documents by most recent change with date and collection filteringlist_collections— List all collectionsget_collection_structure— Get document hierarchy within a collectionget_document_id_from_title— Find document ID by title searchread_document— Get document content with optional line-range paginationexport_document— Export document as markdownget_document_toc— Get table of contents with heading structure and line numbersread_document_section— Read a specific section by heading matchcreate_document— Create new document with optional parent and publish settingsupdate_document— Replace or append to full document contentedit_document— String-match editing with batched replacements and staged changesmove_document— Move document to different collection or parentarchive_document— Archive documentunarchive_document— Restore document from archivedelete_document— Delete document or move to trashrestore_document— Restore document from trashlist_archived_documents— List all archived documentslist_trash— List all documents in trashadd_comment— Add comment to document with threaded reply support
Use cases
- Search your Outline knowledge base and retrieve specific documents by keyword or title
- Create and edit documents programmatically, with support for staged edits and batch operations
- Organize documents by moving them between collections and managing hierarchies
- Ask natural language questions about your document collection using AI-powered search
- Manage document lifecycle including archiving, deletion, and restoration with fine-grained access control
Outline MCP Server MCP server FAQ
It's a Model Context Protocol server that connects AI assistants to Outline document management systems, enabling search, reading, creation, editing, and management of documents and collections via natural language.
Yes, the mcp-outline server is open-source under the MIT License. You need an Outline account (cloud or self-hosted) and an API key from your Outline instance.
Use the one-click install buttons in the README for Cursor or VS Code, or manually add the server config to your MCP client settings with your Outline API key and optional API URL for self-hosted instances.
You need an Outline API key from your Outline web UI (Settings → API Keys → Create New). For self-hosted Outline, also provide the OUTLINE_API_URL environment variable.
Yes, set OUTLINE_READ_ONLY=true to disable all write operations, or OUTLINE_DISABLE_DELETE=true to disable only deletions. You can also use OUTLINE_DYNAMIC_TOOL_LIST=true to filter tools per-user by role and API key scopes.
Yes, set the OUTLINE_API_URL environment variable to your self-hosted instance URL (e.g., https://your-domain/api). You can also disable SSL verification with OUTLINE_VERIFY_SSL=false for self-signed certificates.
README (reference)
Source of truth, from the repository.
MCP Outline Server
📢 Official Outline MCP Server Available
Outline now ships an official MCP server — we recommend using it. Read the docs.
<!-- mcp-name: io.github.Vortiago/mcp-outline -->
A Model Context Protocol server for interacting with Outline document management.
Features
- Document operations: Search, read, create, edit, archive documents
- Collections: List, create, manage document hierarchies
- Comments: Add and view threaded comments
- Backlinks: Find documents referencing a specific document
- MCP Resources: Direct content access via URIs (outline://document/{id}, outline://collection/{id}, etc.)
- Automatic rate limiting: Transparent handling of API limits with retry logic
Prerequisites
Before using this MCP server, you need:
- An Outline account (cloud hosted or self-hosted)
- API key from Outline web UI: Settings → API Keys → Create New
- Python 3.10+ (for non-Docker installations)
Getting your API key: Log into Outline → Click your profile → Settings → API Keys → "New API Key". Copy the generated token.
Quick Start
One-Click Install
Click a button to install with interactive API key prompt:
Manual Install
Install with uv (recommended), pip, or Docker:
uvx mcp-outline # using uv
pip install mcp-outline # using pip
# using Docker
docker run -e OUTLINE_API_KEY=<your-key> ghcr.io/vortiago/mcp-outline:latest
Then add to your MCP client config (works with VS Code, Claude Desktop, Cursor, and others):
{
"inputs": [
{
"id": "outline_api_key",
"type": "promptString",
"description": "Enter OUTLINE_API_KEY",
"password": true
},
{
"id": "outline_api_url",
"type": "promptString",
"description": "Outline API URL (optional, for self-hosted)",
"password": false
}
],
"servers": {
"mcp-outline": {
"command": "uvx",
"args": ["mcp-outline"],
"env": {
"OUTLINE_API_KEY": "${input:outline_api_key}",
"OUTLINE_API_URL": "${input:outline_api_url}"
}
}
}
}
<details>
<summary>Claude Code</summary>
claude mcp add mcp-outline uvx mcp-outline
Installing the repo as a plugin instead also bundles the
outline-explorer agent (fast read-only wiki exploration) and the
outline skill (Outline conventions: mermaidjs fences, document
structure, editing workflows).
Add to ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"mcp-outline": {
"command": "uvx",
"args": ["mcp-outline"],
"env": {
"OUTLINE_API_KEY": "<YOUR_API_KEY>",
"OUTLINE_API_URL": "<YOUR_OUTLINE_URL>"
}
}
}
}
</details>
Setup guides for more clients: Docker (HTTP), Cline, Codex, Windsurf, and others
Configuration
| Variable | Required | Default | Notes |
|---|---|---|---|
OUTLINE_API_KEY | Yes* | - | Required for tool calls to succeed. For SSE/HTTP, can alternatively be provided per-request via x-outline-api-key header (details) |
OUTLINE_API_URL | No | https://app.getoutline.com/api | For self-hosted: https://your-domain/api |
OUTLINE_VERIFY_SSL | No | true | Set false for self-signed certificates |
OUTLINE_READ_ONLY | No | false | true = disable ALL write operations (details) |
OUTLINE_DISABLE_DELETE | No | false | true = disable only delete operations (details) |
OUTLINE_DISABLE_AI_TOOLS | No | false | true = disable AI tools (for Outline instances without OpenAI) |
OUTLINE_DISABLE_RECENT_DOCUMENTS | No | false | true = disable the list_recently_updated_documents tool |
OUTLINE_DYNAMIC_TOOL_LIST | No | false | true = enable per-user tool filtering by role/key scopes (details) |
OUTLINE_MAX_CONNECTIONS | No | 100 | Max concurrent connections in pool |
OUTLINE_MAX_KEEPALIVE | No | 20 | Max idle connections in pool |
OUTLINE_TIMEOUT | No | 30.0 | Read timeout in seconds |
OUTLINE_CONNECT_TIMEOUT | No | 5.0 | Connection timeout in seconds |
OUTLINE_WRITE_TIMEOUT | No | 30.0 | Write timeout in seconds |
OUTLINE_CACHE_TTL | No | 30 | Document cache TTL in seconds. The short default absorbs same-task read bursts without stressing the Outline API; set 0 to disable caching (always-fresh reads) or higher (e.g. 300) for more API savings. Staged edits work either way |
OUTLINE_CACHE_MAX_SIZE | No | 100 | Max cached documents |
MCP_TRANSPORT | No | stdio | Transport mode: stdio (local), sse or streamable-http (remote) |
MCP_HOST | No | 127.0.0.1 | Server host. Use 0.0.0.0 in Docker for external connections |
MCP_PORT | No | 3000 | HTTP server port (only for sse and streamable-http modes) |
Access Control
| Feature | Env Var | Effect |
|---|---|---|
| Read-only mode | OUTLINE_READ_ONLY=true | Disables all write operations — only search, read, and export tools available |
| Disable deletes | OUTLINE_DISABLE_DELETE=true | Disables only delete operations, all other writes allowed |
| Disable recent-changes tool | OUTLINE_DISABLE_RECENT_DOCUMENTS=true | Disables only the list_recently_updated_documents tool |
| Dynamic tool list | OUTLINE_DYNAMIC_TOOL_LIST=true | Filters tools per-user based on Outline role and API key scopes |
| Per-user Outline API keys | x-outline-api-key header | Each user passes their own Outline API key in HTTP mode for multi-user setups |
Read-only mode takes precedence over disable-delete. See Configuration Guide for details.
Tools
Note: Tool availability depends on your access control settings.
Search & Discovery
search_documents(query, collection_id?, limit?, offset?, statusFilter?)- Search documents by keywords with pagination. Defaults to published documents; passstatusFilterwithdraft,archived, and/orpublishedto include other stateslist_recently_updated_documents(date_filter?, collection_id?, status_filter?, limit?, offset?)- List documents by most recent change, newest first (e.g. "what changed this week").date_filterwindows by last change:day/week/month/year(defaultweek). Defaults to published documentslist_collections()- List all collectionsget_collection_structure(collection_id)- Get document hierarchy within a collectionget_document_id_from_title(query, collection_id?)- Find document ID by title search
Document Reading
read_document(document_id, offset?, limit?)- Get document content with optional line-range paginationexport_document(document_id)- Export document as markdown
Document Navigation
get_document_toc(document_id)- Get table of contents with heading structure and line numbersread_document_section(document_id, heading)- Read a specific section by heading match (case-insensitive substring)
Document Management
create_document(title, collection_id, text?, parent_document_id?, publish?)- Create new documentupdate_document(document_id, title?, text?, append?)- Replace full document content (append mode available)
Document Editing
edit_document(document_id, edits, save?)- String-match editing with batched replacements;save=Falsestages changes locally,save=Trueon the final call pushes all changesmove_document(document_id, collection_id?, parent_document_id?)- Move document to different collection or parent
Document Lifecycle
archive_document(document_id)- Archive documentunarchive_document(document_id)- Restore document from archivedelete_document(document_id, permanent?)- Delete document (or move to trash)restore_document(document_id)- Restore document from trashlist_archived_documents()- List all archived documentslist_trash()- List all documents in trash
Comments & Collaboration
add_comment(document_id, text, parent_comment_id?)- Add comment to document (supports threaded replies)list_document_comments(document_id, include_anchor_text?, limit?, offset?)- View document comments with paginationget_comment(comment_id, include_anchor_text?)- Get specific comment detailsget_document_backlinks(document_id)- Find documents that link to this document
Collection Management
create_collection(name, description?, color?)- Create new collectionupdate_collection(collection_id, name?, description?, color?)- Update collection propertiesdelete_collection(collection_id)- Delete collectionexport_collection(collection_id, format?)- Export collection (default: outline-markdown)export_all_collections(format?)- Export all collections
Batch Operations
batch_create_documents(documents)- Create multiple documents at oncebatch_update_documents(updates)- Update multiple documents at oncebatch_move_documents(document_ids, collection_id?, parent_document_id?)- Move multiple documentsbatch_archive_documents(document_ids)- Archive multiple documentsbatch_delete_documents(document_ids, permanent?)- Delete multiple documents
AI-Powered
ask_ai_about_documents(question, collection_id?, document_id?)- Ask natural language questions about your documents
Resources
outline://collection/{id}- Collection metadata (name, description, color, document count)outline://collection/{id}/tree- Hierarchical document tree structureoutline://collection/{id}/documents- Flat list of documents in collectionoutline://document/{id}- Full document content (markdown)outline://document/{id}/backlinks- Documents that link to this document
Development
git clone https://github.com/Vortiago/mcp-outline.git
cd mcp-outline
uv sync --group dev
uv run poe test-unit # unit tests
uv run poe test-integration # integration tests (starts MCP server via stdio)
uv run poe test-e2e # E2E tests (requires Docker)
See Development Guide for self-hosted Outline setup, MCP Inspector, and more.
Troubleshooting
Server not connecting? Test your API key:
curl -H "Authorization: Bearer YOUR_API_KEY" YOUR_OUTLINE_URL/api/auth.info
See Troubleshooting Guide for common issues with tools, rate limiting, and Docker.
Contributing
Contributions welcome! See CONTRIBUTING.md for setup instructions.
License
This project is licensed under the MIT License - see the LICENSE file for details.
Acknowledgments
- Built with MCP Python SDK
- Uses Outline API for document management
Related MCP servers
Shared memory for AI agents. One address per fact, one signature you check. No key to read.

io.github.VouchlyAI/pincer
Secure grip for your agent's secrets - security-hardened MCP gateway with proxy token architecture

Publish AI images & carousels to Vynly, the AI-only social feed. Verified provenance, no signup.
Local-first MCP tools to create, validate, and preview OpenFlowKit diagrams from Claude, Cursor, or Windsurf.

io.github.Vublox/sports
Live football scores, match summaries, and fan clip links from Vublox. No API key required.


