mcp-registry/zettelkasten-mcp

    ==================
      
       /// MCP ///
      /// ZET ///
        
    ==================
        
    [server:online]
    [protocol:ready]

zettelkasten-mcp

by entanglr

Model-Context-Protocol server implementing the Zettelkasten knowledge-management method; provides CLI tools and HTTP interface to create, link and query atomic Markdown notes backed by SQLite.

69
12
Specialized
GitHubDocumentation

Installation

1. Prerequisites
• Python ≥3.9
• A working Anthropic (Claude) API key or other MCP-compatible LLM key in your environment (e.g. ANTHROPIC_API_KEY)
• Git (optional if you install from PyPI once it’s published)
2. Clone and enter the project directory
git clone https://github.com/entanglr/zettelkasten-mcp.git cd zettelkasten-mcp
3. Create and activate a virtual environment (recommended)
python -m venv .venv source .venv/bin/activate # Windows: .venv\Scripts\activate
4. Install server and dependencies
pip install -r requirements.txt # production deps pip install -r requirements-dev.txt # optional: tests & linting
(If the project is published on PyPI you can alternatively run `pip install zettelkasten-mcp`.)
5. Configure runtime settings (env vars or .env file)
ANTHROPIC_API_KEY="<your_claude_key>" ZK_ROOT="/absolute/path/to/your/notes" # folder containing markdown files ZK_INDEX_FILE="index.json" # auto-generated index MCP_HOST="0.0.0.0" MCP_PORT=8787
6. Launch the MCP server
python -m zettelkasten_mcp.run # or: uvicorn zettelkasten_mcp.api:app --host $MCP_HOST --port $MCP_PORT
7. Test the endpoint
Visit `http://localhost:8787/mcp/v1/status` in your browser or run:
curl http://localhost:8787/mcp/v1/status
8. Connect from an MCP client (e.g. Claude, Obsidian plugin, Postman) by pointing it at `http://<server>/mcp/`.

Documentation

License: MIT License
Updated 7/30/2025