anytype
An MCP server enabling AI assistants to interact with Anytype - your encrypted, local and collaborative wiki - to organize objects, lists, and more through natural language.
claude mcp add --transport stdio anyproto-anytype-mcp npx -y @anyproto/anytype-mcp \
--env OPENAPI_MCP_HEADERS="{"Authorization":"Bearer <YOUR_API_KEY>", "Anytype-Version":"2025-11-08"}"How to use
This MCP server exposes Anytype's API as a set of MCP tools, enabling AI assistants to interact with your Anytype workspace through natural language. It translates Anytype's OpenAPI specification into MCP actions such as global and space searches, managing spaces and members, handling objects and lists, and working with properties, tags, types, and templates. To get started, provide your Anytype API key and run the MCP server via a client like Claude, Cursor, Windsurf, or Raycast. Once running, you can issue natural language requests such as creating spaces, adding objects, or performing searches, and the MCP server will convert them into the appropriate MCP tool calls against Anytype.
Key capabilities include: Global & Space Search, managing Spaces & Members, Objects & Lists, Properties & Tags, and handling Types & Templates. Use the configured OPENAPI_MCP_HEADERS with your API key to authenticate requests, and, if needed, set ANYTYPE_API_BASE_URL to point to a custom base URL (default is http://127.0.0.1:31009). The server is designed to work as a drop-in MCP backend for common clients, so you can reuse existing MCP prompts and workflows to interact with your knowledge base in Anytype.
How to install
Prerequisites:
- Node.js (LTS version, e.g., 18.x) and npm installed on your machine
- An Anytype API key (generate from Anytype App Settings > API Keys)
- Optional: a preferred MCP client (Claude Desktop, Cursor, Windsurf, Raycast, etc.)
Installation steps (per-project):
- Initialize your project (if needed) and install the MCP package locally:
npm init -y
npm install @anyproto/anytype-mcp
- Run the MCP server using npx (as shown in the Quick Start):
npx -y @anyproto/anytype-mcp
- If you want to use a global installation instead:
npm install -g @anyproto/anytype-mcp
Then you can run the server via:
anytype-mcp
- Configure your MCP client (example shown in the README):
{
"mcpServers": {
"anytype": {
"command": "npx",
"args": ["-y", "@anyproto/anytype-mcp"],
"env": {
"OPENAPI_MCP_HEADERS": "{\"Authorization\":\"Bearer <YOUR_API_KEY>\", \"Anytype-Version\":\"2025-11-08\"}"
}
}
}
}
Prerequisites recap: ensure your environment has Node.js and npm installed, obtain an Anytype API key, and choose a compatible MCP client to integrate the anytype-mcp server into your workflow.
Additional notes
Tips and common considerations:
- Always supply OPENAPI_MCP_HEADERS with your API key and Anytype-Version to authenticate requests. Use a secure method to store and reference this key in your environment.
- If you run approaches against a non-default API base URL (ANYTYPE_API_BASE_URL), ensure the port and path match your Anytype instance (e.g., http://localhost:31012 for anytype-cli).
- If you prefer a global installation, you can reference the command anytype-mcp in your MCP client configuration; otherwise, using npx ensures you’re running the package version from npm.
- When testing, start with simple interactions (e.g., search or create a small object) to verify authentication and base URL configuration before expanding to more complex prompts.
- If you encounter version compatibility issues, verify you’re using a supported Node.js version and that your MCP client payloads align with the OPENAPI_MCP_HEADERS expectations.
Related MCP Servers
mcp-graphql
Model Context Protocol server for GraphQL
tableau
Tableau's official MCP Server. Helping Agents see and understand data.
filesystem
A Model Context Protocol (MCP) server for platform-agnostic file capabilities, including advanced search/replace and directory tree traversal
any-script
An MCP server that exposes arbitrary CLI tools and shell scripts as MCP Tools
postman
A Model Context Protocol (MCP) server that provides seamless integration with the Postman API. This package enables AI assistants and applications to interact with Postman workspaces, collections, requests, environments, and folders programmatically.
google -remote
Collection of Google-native tools (e.g., Gmail, Calendar) for the MCP