outline-mcp-server

6am/outline-mcp-server

3.1

If you are the rightful owner of outline-mcp-server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.

A Model Context Protocol (MCP) server for Outline that facilitates document management through the Outline API.

Tools
8
Resources
0
Prompts
0

Outline MCP Server

A Model Context Protocol (MCP) server for Outline that enables reading and writing documents through the Outline API.

Features

  • Read Documents: Get individual documents, search, and list documents
  • Write Documents: Create, update, and delete documents
  • Collection Management: List and retrieve collection information
  • Full Text Search: Search across all documents in your Outline instance
  • Markdown Support: Create and edit documents with full Markdown formatting

Installation

  1. Clone or download this repository
  2. Install dependencies: 
    npm install
  3. Build the project: 
    npm run build

Configuration

Before using the server, you need to set up your Outline API credentials:

  1. Get your Outline API token:

  2. Set environment variables:

    export OUTLINE_BASE_URL="https://your-outline-instance.com"
export OUTLINE_API_KEY="your-api-token-here"

Usage

Running the Server

Start the MCP server:

npm start

The server communicates via stdio and is compatible with any MCP client.

Available Tools

Document Operations

  1. outline_get_document

    • Get a specific document by ID
    • Parameters: id (string, required)

  2. outline_search_documents

    • Search for documents across your Outline instance
    • Parameters: query (string, required), limit (number, optional, default: 25)

  3. outline_list_documents

    • List documents, optionally filtered by collection
    • Parameters: collectionId (string, optional), limit (number, optional, default: 25)

  4. outline_create_document

    • Create a new document
    • Parameters:
      • title (string, required)
      • text (string, required) - Markdown content
      • collectionId (string, optional)
      • parentDocumentId (string, optional)
      • publish (boolean, optional, default: false)

  5. outline_update_document

    • Update an existing document
    • Parameters:
      • id (string, required)
      • title (string, optional)
      • text (string, optional) - Markdown content
      • publish (boolean, optional)

  6. outline_delete_document

    • Delete a document
    • Parameters: id (string, required)
Collection Operations

  1. outline_list_collections

    • List all collections in your Outline instance
    • Parameters: none

  2. outline_get_collection

    • Get information about a specific collection
    • Parameters: id (string, required)

Example Usage

Here are some example tool calls:

{
  "name": "outline_search_documents",
  "arguments": {
    "query": "project documentation",
    "limit": 10
  }
}

{
  "name": "outline_create_document",
  "arguments": {
    "title": "New Project Plan",
    "text": "# Project Overview\n\nThis document outlines...",
    "collectionId": "collection-id-here",
    "publish": true
  }
}

{
  "name": "outline_update_document",
  "arguments": {
    "id": "document-id-here",
    "title": "Updated Project Plan",
    "text": "# Updated Project Overview\n\nThis document has been updated..."
  }
}

Development

Project Structure

src/
├── index.ts           # Main MCP server implementation
├── outline-client.ts  # Outline API client

Scripts

  • npm run build - Compile TypeScript to JavaScript
  • npm run dev - Build and run the server
  • npm run watch - Watch for changes and rebuild
  • npm start - Run the compiled server

Building

npm run build

The compiled JavaScript will be output to the dist/ directory.

Configuration with MCP Clients

To use this server with an MCP client, you'll need to configure it to run this server. The exact configuration depends on your client, but generally you'll need to:

  1. Specify the command to run: node /path/to/outline-mcp-server/dist/index.js
  2. Set the environment variables for your Outline instance
  3. Configure the client to use stdio transport

Client Configuration Examples

Claude

For clients like Claude that use a JSON configuration file, you can add the following to your mcp-servers.json:

{
  "mcpServers": {
    "outline": {
      "command": "node",
      "args": ["/path/to/your/projects/outline-mcp-server/dist/index.js"],
      "env": {
        "OUTLINE_API_KEY": "your-secret-api-token",
        "OUTLINE_BASE_URL": "https://your-outline-instance.com"
      }
    }
  }
}

Make sure to replace the args path with the absolute path to the index.js file in your project, and fill in your actual credentials in the env block.

Cursor

For clients like Cursor, you can typically set environment variables directly within the client's settings or by launching the client from a terminal where you have already exported the variables.

export OUTLINE_BASE_URL="https://your-outline-instance.com"
export OUTLINE_API_KEY="your-secret-api-token"

# Then launch Cursor from this terminal
/path/to/Cursor.app/Contents/MacOS/Cursor

API Rate Limits

Be aware that Outline may have API rate limits. The server doesn't implement rate limiting internally, so you may need to handle this at the client level if you're making many requests.

Error Handling

The server includes comprehensive error handling and will return descriptive error messages for common issues like:

  • Missing or invalid API credentials
  • Network connectivity problems
  • Invalid document IDs
  • API rate limit errors

Security Notes

  • Store your API token securely using environment variables
  • Never commit your API token to version control
  • Consider using restricted API tokens with minimal required permissions
  • Be cautious when allowing others to use your MCP server as it has full access to your Outline instance

License

MIT License - see LICENSE file for details.

Using with Different Platforms

ChatGPT (GPT Store)

To use this MCP server with ChatGPT:

  1. Install via npm (when published):

    npm install -g outline-mcp-server

  2. Set up as a connector:

    • In ChatGPT, go to Settings → Connectors
    • Add new connector with URL: https://your-domain.com/mcp-endpoint
    • Configure OAuth or API key authentication

  3. Alternative: Use with ChatGPT Actions:

    • Create custom GPT Actions pointing to your hosted MCP server
    • Define the OpenAPI schema based on the MCP tools

n8n Integration

To use with n8n:

  1. HTTP Request Node: Configure n8n to make HTTP requests to your MCP server endpoint
  2. Custom Node: Create a custom n8n node that wraps the MCP functionality
  3. Webhook: Set up webhooks to trigger document operations

Example n8n workflow:

{
  "nodes": [
    {
      "name": "Outline Search",
      "type": "n8n-nodes-base.httpRequest",
      "parameters": {
        "url": "http://localhost:3000/tools/outline_search_documents",
        "method": "POST",
        "body": {
          "query": "{{$json.searchTerm}}"
        }
      }
    }
  ]
}

Make.com Integration

Similar to n8n, you can use HTTP modules to interact with the MCP server:

  1. Create a scenario with HTTP Request modules
  2. Configure authentication with your Outline credentials
  3. Map the MCP tool calls to HTTP requests

General REST API Usage

You can also wrap this MCP server with a REST API:

import express from 'express';
import { OutlineClient } from './outline-client.js';

const app = express();
app.use(express.json());

const client = new OutlineClient();

app.post('/search', async (req, res) => {
  const { query, limit } = req.body;
  const results = await client.searchDocuments({ query, limit });
  res.json(results);
});

app.listen(3000);

NPM Publication

To publish this package to npm for easier installation:

  1. Update package.json:

    {
  "name": "@your-username/outline-mcp-server",
  "version": "1.0.0",
  "publishConfig": {
    "access": "public"
  }
}

  2. Login to npm:

    npm login

  3. Publish:

    npm publish

  4. Install globally:

    npm install -g @your-username/outline-mcp-server

Contributing

Contributions are welcome! Please feel free to submit issues and pull requests.