
Shelbymcp
Gives Claude and other MCP-compatible tools a persistent knowledge graph instead of starting every conversation from scratch. Nine tools let you capture thoughts with typed relationships (refines, cites, refuted_by), search via FTS5 or embeddings, and traverse the graph to pull in connected context. Ships with a Forage skill that runs on your own Claude Code subscription to backfill embeddings, deduplicate memories, and detect contradictions over time. Single SQLite file, no Docker or Python runtime. Auto-configures for Claude Desktop, Cursor, Windsurf, Gemini CLI, and others. The memory layer is extracted from Shelby, the Mac AI coworker, so you can run it standalone with whatever editor or CLI you're already using.
- tools
- 6
- transport
- STDIO
- last updated
- Mar 29, 2026
Tools
6 toolsPublic tool metadata for what this MCP can expose to an agent.
manage_edgesLink and traverse typed graph edges.
Link and traverse typed graph edges.
No parameter schema in public metadata yet.
explore_graphLink and traverse typed graph edges.
Link and traverse typed graph edges.
No parameter schema in public metadata yet.
expand_neighborsLink and traverse typed graph edges.
Link and traverse typed graph edges.
No parameter schema in public metadata yet.
get_briefBuild trusted, token-bounded context.
Build trusted, token-bounded context.
No parameter schema in public metadata yet.
select_contextBuild trusted, token-bounded context.
Build trusted, token-bounded context.
No parameter schema in public metadata yet.
thought_statsAudit the database.
Audit the database.
No parameter schema in public metadata yet.
Sponsors
Readme
Local knowledge-graph memory for AI tools, implemented as one Rust binary.
Advisaur MCP gives MCP-compatible agents durable memory across sessions. Thoughts live in a local SQLite database with FTS5 search, optional vectors, typed graph edges, project scope, trust fencing, and curated context briefs. The server performs no inference and makes no model API calls.
Quick start
Run the platform-native binary through npm:
npx -y advisaur-mcp --version
npx -y advisaur-mcp
The default stdio server stores data at ~/.studiomoser-assistant-mcp/memory.db. Existing databases created by Advisaur MCP schema v18 open unchanged.
The server was called ShelbyMCP before its rename (ADR-0012). An existing ~/.shelbymcp folder moves to the new one the first time the server runs with its default database, and the old SHELBY_DB_PATH, SHELBY_TRANSPORT and SHELBY_API_KEY variables are still read, with a warning, for one release. npx shelbymcp still runs the server and says where it went.
Package-first integrations are included in the protected v0.4.0 release:
Details
- Repository
- github.com/studio-moser/shelbymcp
- Categories
- AI & LLM Tools, Documents & Knowledge
- Package
shelbymcp- Transport
- STDIO
- Registry
- MCP registry, active