About this integration
Local-first memory for Claude Code and any MCP client: hybrid search + knowledge graph, $0/token.
- Transport
- stdio
- Authentication
- none
- Initial setup
- none
- Runtime
- unattended
- Evidence
- documented
- Version
- 2.10.3
- Package
- mcp-memory-graph
- Last compatibility test
- Not independently tested
Connect your agent
mcp-memory-graphPublisher documentation reviewed; package not executed. The default local stdio path needs no cloud account or API key; remote HTTP mode has separate bearer-token requirements.
Capabilities: Store agent memory, Search memories semantically, Run local consolidation
Connected profiles
Additional details
io.github.YonasValentin/mcp-memory-graph
Source ↗ · Checked 2026-09-172.10.3
Source ↗ · Checked 2026-09-17CC0-1.0; package licenses are separate
Source ↗ · Checked 2026-09-17npm
Source ↗ · Checked 2026-09-17mcp-memory-graph
Source ↗ · Checked 2026-09-202.10.3
Source ↗ · Checked 2026-09-20The Stop hook detaches in about 30 ms and reviews in the background for 10 to 60 seconds. It needs the `claude` CLI on `$PATH` (or `$CLAUDE_BIN`), authenticated. Turn it off with `review_on_stop: false` in `~/.mcp-memory/config.json`.
Source ↗ · Checked 2026-09-20A memory server for [Claude Code](https://docs.anthropic.com/en/docs/claude-code) and any other [MCP](https://modelcontextprotocol.io/) client. It gives your AI assistant a permanent, searchable memory that lives in one SQLite file on your machine. Store a decision today, ask about it next month, and the answer comes back. Everything runs locally: the embedding model, the search index, the knowledge graph. No cloud account, no API key, no per-token cost.
Source ↗ · Checked 2026-09-20npx mcp-memory-graph init
Source ↗ · Checked 2026-09-20Publisher documentation reviewed; package and integration endpoint not executed or independently security-audited.
Source ↗ · Checked 2026-09-20