Skip to main content
Glama

Vault Brain 🧠

High-performance GraphRAG knowledge graph retrieval engine for Obsidian Vaults. Built for Hermes Agent & Vickry's personal second brain.

Features

  • Deterministic Obsidian Parser: Fast regex & AST extraction of YAML frontmatter, wikilinks ([[Note#Section|Alias]]), headings, and tags.

  • Strict Ontology: Enforces verified entity types (NOTE, CONCEPT, PROJECT, TOOL, PERSON) and relations (RELATES_TO, DEPENDS_ON, IMPLEMENTS, PART_OF, OWNS, MENTIONED_IN).

  • In-Memory NetworkX + SQLite Persistence: Lightning-fast graph traversal and pathfinding (<50ms) backed by persistent disk storage (vault.db).

  • GraphRAG Semantic Retrieval: Multi-hop semantic traversal and context-aware evidence snippet extraction.

  • Native MCP Server: Exposes query_graph, find_related, trace_topic, and vault_hubs directly to Hermes and Claude Code.

  • Zero Heavy External Dependencies: Pure Python (networkx, pyyaml, mcp), RAM footprint < 80MB.

Related MCP server: Obsidian Knowledge MCP

CLI Usage

# Index Obsidian vault
vaultbrain index --vault /root/obsidian-vault --out /root/vaultbrain/vault.db

# Query graph for a concept or keyword
vaultbrain query "FastAPI"

# Trace topic evolution across notes
vaultbrain trace "Beemate"

# Find paths between two notes
vaultbrain path "Pikri" "FastAPI" --max-hops 3

Running Tests

cd /root/vaultbrain
/usr/local/lib/hermes-agent/venv/bin/pytest tests/ -v

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      A
      maintenance
      Connects AI assistants to an Obsidian vault as a semantic knowledge graph, enabling graph navigation, semantic search, and content operations through MCP.
      471
      MIT
    • F
      license
      Not graded
      quality
      D
      maintenance
      Enables read-only querying of an Obsidian knowledge base through MCP tools and resources, including full-text search, node and link retrieval, related-note traversal, and pathfinding between concepts.
      1
      -
    • A
      license
      A
      quality
      B
      maintenance
      Exposes a typed, provenance-aware knowledge graph over an Obsidian vault to coding agents over MCP, so they can trace what grounds a claim, compute which other notes a change makes wrong, surface contradictions and admitted gaps, and assemble a task-specific token-budgeted subset of the vault. Indexing, retrieval, and impact analysis run with zero third-party runtime dependencies beyond the Python standard library.
      10
      MIT
    • A
      license
      B
      quality
      B
      maintenance
      Enables AI agents and developers to inspect Obsidian vault knowledge graphs, navigate backlinks, detect broken wikilinks, lint note compliance, search indexed notes, and scaffold standards-compliant Markdown notes via MCP.
      6
      MIT