Markdown Vault MCP
Markdown Vault MCP is a generic MCP server that exposes a folder of Markdown notes, such as an Obsidian vault, to an AI client with hybrid keyword and semantic search, frontmatter-aware filtering, and optional write tools for editing the vault directly.
Search combines SQLite FTS5 keyword search with semantic search from FastEmbed, Ollama, OpenAI, Voyage AI or any OpenAI-compatible endpoint, fused with reciprocal rank fusion and returned as diversity-aware sentence-scale snippets that can expand to a full section on request. YAML frontmatter fields become filterable and searchable automatically, and long documents are chunked adaptively by heading level.
Write tools such as write, edit, append, delete, rename and move are registered by default but can be hidden entirely by setting a read-only environment variable, and every write updates the search index automatically; an optional Git integration can auto-commit each write, push on a delay, and pull external changes through a webhook, with revision history letting an overwritten note be read back at a prior commit.
It is written in Python, distributed on PyPI and as a Docker image, and released under the MIT license.