metamind-vault-rag
MCP server for hybrid retrieval over markdown vaults, combining sqlite-vec embeddings and FTS5 keywords with reciprocal rank fusion, plus optional cross-encoder reranking.
README
metamind-vault-rag
A retrieval engine for a directory of markdown. It watches files, indexes them incrementally, and answers hybrid search queries over them.
Vectors live in sqlite-vec, keywords in SQLite FTS5, and the two are fused with reciprocal rank fusion. Embeddings run in-process through fastembed's ONNX models, so there is no server to stand up and no API key to hold. An optional cross-encoder rescore tier is available through the rerank extra.
Install
uv tool install metamind-vault-rag
Entry points
| Command | Purpose |
|---|---|
metamind-vault-rag-watcher |
Watch a directory and index changes |
metamind-vault-rag-indexer |
One-shot full reindex |
metamind-vault-rag-http |
Loopback HTTP search API |
metamind-vault-rag-server |
stdio MCP server |
metamind-vault-rag-doctor |
Environment and index diagnostics |
Configuration
| Variable | Meaning |
|---|---|
VAULT_PATH |
Directory to index |
VAULT_COLLECTION |
Collection name, which scopes the index files |
VAULT_HTTP_PORT |
Port for the loopback search API |
VAULT_STATE_DIR |
Where indexes, caches and logs are written. Defaults to ~/.vault-rag |
Indexes are written to the state directory, named after the collection, and are never placed inside the corpus. Two clients pointed at different collections, or different state directories, coexist on one machine without either knowing about the other.
Consumers
Installed by any client that wants retrieval without running a service. The engine holds no opinion about who is asking: it names no client in its output, its environment variables are all VAULT_-prefixed, and it writes nothing outside the state directory.
Development
uv run --extra dev pytest
A client can be pointed at a working copy instead of a release with uv tool install --from /path/to/this/repo metamind-vault-rag.
Licence
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.