MarkusPfundstein/mcp-obsidian

What it solves

This project provides a bridge between AI assistants (like Claude) and your personal knowledge base in Obsidian. It allows an LLM to read, search, and edit your notes, effectively turning your vault into a dynamic memory for the AI.

How it works

The project implements a Model Context Protocol (MCP) server. It connects to the Obsidian Local REST API community plugin, which allows the server to execute specific tools for managing files. The AI assistant uses these tools to interact with your vault via JSON-RPC over stdio.

Who it’s for

Users of Obsidian who want to use AI to summarize notes, search for specific context across their vault, or automatically generate and update content within their notes.

Highlights

• Vault Management: Supports listing files, reading contents, and deleting files.
• Content Manipulation: Ability to append content or patch content relative to headings, block references, or frontmatter.
• Search: Full-text search across all files in the vault.
• Deployment Options: Can be run via uvx, Python locally, or as a Docker container.