affine-mcp-server

Overview

The affine-mcp-server is a well-structured MCP server that enables AI assistants to interact with AFFiNE workspaces. It provides comprehensive access to documents, databases, collaboration features, and organizational tools through a stable API surface. The server supports both stdio and HTTP transports, with robust authentication options including tokens, cookies, and email/password for both AFFiNE Cloud and self-hosted instances.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• create_workspace

  Create a new workspace in AFFiNE

  Try:Create a new workspace for my projectTry:Set up a workspace called 'Meeting Notes'
• search_documents

  Search for documents across workspaces

  Try:Find all documents about project planningTry:Search for documents containing the term 'Q3 results'
• read_document

  Read the content of a specific document

  Try:Show me the content of document ID 123Try:Read the document titled 'Annual Report'
• create_document

  Create a new document in a workspace

  Try:Create a new document called 'Project Ideas'Try:Start a new meeting notes document
• create_database_column

  Add a new column to a database

  Try:Add a 'Priority' column to the tasks databaseTry:Create a 'Due Date' column in my project tracker
• add_database_row

  Add a new row to a database

  Try:Add a new task to the databaseTry:Create a row for the project milestone in the tracker
• list_comments

  List all comments on a document

  Try:Show me all comments on the design documentTry:List comments on the budget proposal
• create_comment

  Add a comment to a document

  Try:Add a comment about the timelineTry:Comment on the section about budget
• get_user_info

  Retrieve information about the current user

  Try:Show my user profileTry:Get my account information
• upload_blob

  Upload a file/blob to AFFiNE storage

  Try:Upload the presentation PDFTry:Add the design mockup to the workspace
• list_notifications

  Get all notifications for the current user

  Try:Show me my notificationsTry:List all unread notifications
• publish_document

  Publish a document to make it publicly accessible

  Try:Publish the company handbookTry:Make the project documentation public

Note: Tool names inferred from the Tool Surface section which lists domains but doesn't provide explicit tool names. The actual tool names would be defined in tool-manifest.json.

Comparable tools

Installation

npm i -g affine-mcp-server
affine-mcp --version

npx -y -p affine-mcp-server affine-mcp -- --version

docker run -d 
  -p 3000:3000 
  -e MCP_TRANSPORT=http 
  -e AFFINE_BASE_URL=https://your-affine-instance.com 
  -e AFFINE_API_TOKEN=ut_your_token 
  -e AFFINE_MCP_AUTH_MODE=bearer 
  -e AFFINE_MCP_HTTP_TOKEN=your-strong-secret 
  ghcr.io/dawncr0w/affine-mcp-server:latest

{
  "mcpServers": {
    "affine": {
      "command": "affine-mcp"
    }
  }
}

FAQ

What authentication methods are supported?

The server supports token, cookie, and email/password authentication for both AFFiNE Cloud and self-hosted instances. API tokens are recommended for production use.

Can I restrict which tools are exposed?

Yes, you can use environment variables like AFFINE_TOOL_PROFILE (read_only, core, authoring) or AFFINE_DISABLED_GROUPS to limit tool exposure for least-privilege deployments.

Compare affine-mcp-server with