Notemd

You can run the Notemd MCP Server to offload heavy AI tasks for your Markdown vaults. This server exposes endpoints for content enrichment, documentation generation, web research, and automatic backlink management, empowering your knowledge graph with fast, centralized processing.

How to use

Connect an MCP client to the Notemd MCP Server to access its AI-powered capabilities. Use the standard versus custom endpoints to enrich content, generate documentation, perform web research and summarize results, or manage backlinks when files move or disappear. You will interact with the server through your MCP client by selecting the Notemd MCP Server as a provider and invoking the available operations. Begin by starting the server via your chosen method, then configure your client to point at the local host and the server’s API surface.

How to install

Follow these concrete, step-by-step instructions to run the Notemd MCP Server on your machine.

# Prerequisites
# Choose one of the two execution paths below
# Path A: Using npx (quick start)
npx notemd-mcp-server

# Path B: Local installation with uvx or pip
# 1) Install Python dependencies (uvx path)
uv venv
uv pip install -r requirements.txt

# 2) Run the server (uvx path)
uvicorn main:app --reload

MCP configuration (example)

If you are wiring this server into a Mission Control Platform (MCP) configuration, you can reference an MCP entry that starts the server via npx. The following example shows how to declare the MCP entry and required environment variables.

{
  "mcpServers": {
    "notemd_mcp": {
      "description": "Notemd MCP Server - AI-powered text processing and knowledge management for Markdown files.",
      "command": "npx",
      "args": ["notemd-mcp-server"],
      "env": {
        "OPENAI_API_KEY": "your_openai_api_key_here",
        "DEEPSEEK_API_KEY": "your_deepseek_api_key_here"
      }
    }
  }
}

Available tools

process_content

Enriches a block of text by adding wiki-links, building a connected knowledge graph from your Markdown content.

generate_title

Creates comprehensive documentation from a single title or keyword, optionally including context from web research.

research_summarize

Performs a web search and returns an AI-generated concise summary of the topic.

execute_custom_prompt

Executes a user-defined prompt against provided content, returning the AI-generated response.

handle_file_rename

Updates backlinks across the vault when a file is renamed to preserve link integrity.

handle_file_delete

Removes backlinks to a file that has been deleted to prevent broken links.

batch_fix_mermaid

Scans Markdown files in a folder and corrects common Mermaid.js and LaTeX syntax errors.

health

Health check endpoint to verify the MCP server is running.