brain-mcp

brain-mcp

Your Obsidian vault, as memory for every AI you use.

An open-source Model Context Protocol server that serves your notes, from your own machine, to Claude and any MCP client.

Website · Quick start · How it works · Contribute · Research questions

---

Why brain-mcp

Every new chat starts from zero. You explain who you are, what you are building and how you like to work, and tomorrow, in another app, you explain it again.

brain-mcp keeps that context in the Markdown notes you already own. Connect any MCP client and it learns who you are, how you work, what your projects are and what you touched this week. It can search, follow links and write back to the vault as you talk.

• Your files, your machine. Notes stay plain Markdown in a folder you control. Nothing is uploaded to a third party.
• One memory, every client. Claude Code, Claude Desktop, claude.ai, Cursor, Hermes Agent, your phone and anything else that speaks MCP read the same vault.
• Hybrid retrieval, fully local. Keyword search plus meaning search with a small embedding model, then a reranker, all on your CPU. It can also answer "the vault does not record this" instead of guessing.
• Measured. Every ranking change is scored by two bundled evals, keyword and hybrid side by side.
• Private by default. OAuth 2.1 with PKCE for remote access, read, write and private scopes, secret redaction and an audit log of every call.
• Writes back. brain_remember and brain_capture turn what a model learns into notes you can read and edit.
• Small and readable. About 3,800 lines of strict TypeScript. Easy to study, easy to extend.

How it works

flowchart LR

V["Obsidian vault plain Markdown"]

subgraph S["brain-mcp: one Node process, one SQLite file"] I["Indexer sections, links, topics, projects"] DB[("SQLite FTS5 word index section embeddings link graph")] M["Local models on CPU bge-small embedder bge-reranker"] R["Hybrid retrieval"] T["14 MCP tools"] G["Scopes, redaction, audit log"] I --> DB --> R --> T --> G M -.-> DB M -.-> R end

V -->|file watcher| I G -->|stdio| C1["Claude Code, Claude Desktop"] G -->|HTTP + bearer token| C2["Clients on your network"] G -->|OAuth 2.1 through Cloudflare Tunnel| C3["claude.ai, phone, any MCP client"] T -. "remember, capture, write, edit, move, delete (allowed folders only)" .-> V

1. Index. Every note is split into sections and indexed with its links, topics and project. Each section also gets an embedding (bge-small-en-v1.5) in the background. A file watcher keeps both current within a second of a save.
2. Retrieve. A question goes through five steps:

• Understand: expand shorthands (pg → PostgreSQL) and fix typos against the vault's own words.
• Search four ways: keywords per section (BM25), meaning (embeddings), exact dates, and note titles.
• Fuse the four lists with reciprocal rank fusion.
• Follow links (off by default, retrieval.graph): a section of a note linked with the top hits (hub and topic pages included, with a boost when several top hits link to the same note) can take a reranker place from the weakest fused entry it beats, and the reranker reads the sentence that links them. A section the last step would trim is kept anyway when it is linked with a kept hit.
• Rerank the best candidates with a cross-encoder (bge-reranker-base), always including each list's first pick, so fusion cannot bury a hit only one search found. When keyword and meaning search agree on the top hit and fusion keeps it in the top 3, only those 3 are reranked.
• Decide: if even the best passage scores below the reranker's floor, answer "nothing relevant"; otherwise return the strong hits and at least the five best, cited by note, with a coverage label (good, thin, none).

1. Serve. Over stdio beside your editor, over HTTP with a bearer token on your network, or behind an OAuth 2.1 login through Cloudflare Tunnel for the public internet.
2. Remember and write. brain_remember appends durable facts and refuses near duplicates. A fact can replace an earlier one, expire, or start at a given date; brain_identity serves only the facts true now, and brain_facts keeps the history. Four more tools write, edit, move and delete hand-written notes, inside the folders you allow.

The retrieval step from question to answer:

flowchart LR

Q["Question"] --> U["Understand shorthands, typo fixes"] U --> K["Keywords per section, BM25"] U --> D["Meaning embeddings"] U --> DT["Exact dates"] U --> NT["Note titles BM25"] K & D & DT & NT --> F["Fuse reciprocal rank fusion"] F --> RR["Rerank top 16 cross-encoder (top 3 when keywords and meaning agree)"] RR --> GT{"Best score above the floor?"} GT -->|no| X["Nothing relevant"] GT -->|yes| A["Cited passages coverage: good or thin"]

Everything runs on your machine: the models are downloaded once into data/models, and no note leaves the server. If the models are not ready yet, retrieval falls back to keywords, so the server never blocks. Set BRAIN_MCP_HYBRID=0 to stay keyword-only.

[!NOTE]

The architecture is intentionally simple right now, and ideas are welcome. It is one process and one SQLite file, with a hybrid ranker built from fixed rules. Some of it is already tweakable: the retrieval block in brain.config.json (embedding and reranker models, how many candidates to rerank, the score blend, the "nothing relevant" floor, your own shorthands), context budgets under context, and fusion constants such as RRF_K and TITLE_BONUS in src/vault/index.ts. Much more could become configurable, such as pluggable retrievers and storage, graph strategies, and query rewriting. If you have an idea, open an issue or a research proposal, even before there is code.

The bundled example vault as a knowledge graph on the project site . Drag to rotate, hover a node to read the note.

Tools

The four write tools need the brain:write scope. They refuse notes a pipeline generates, hidden folders, paths outside the vault, and the memory file, which stays append-only through brain_remember.

Also exposed: the resources brain://identity and brain://note/{path}, and the prompts assume_persona and project_briefing.

The index lives in data/index.db, embeddings included. It is rebuilt on every start (about 10 ms for the 15-note example vault) and kept current by the watcher; embeddings are keyed by a hash of the text, so unchanged sections are never embedded twice. Every tool call is appended to logs/audit.jsonl.

Quick start

Requires Node 22 or newer. One command sets i