Store and retrieve agentic memories locally with hybrid search, GDPR compliance, and no cloud dependency. Built for MCP agents that need persistent, privacy-first context management across conversations.
io.github.skynetcmd/m3-memory
Local install
STDIO
No auth required
How models use it and what it is built for.
Store and retrieve agentic memories locally with hybrid search, GDPR compliance, and no cloud dependency. Built for MCP agents that need persistent, privacy-first context management across conversations.
Local install — runs as a subprocess.
Where to find authoritative docs and source for M3 Memory.
Paste any of these into Agent Studio after connecting M3 Memory.
Common questions about connecting and running M3 Memory.
What does m3-memory do?
m3-memory is a local-first memory system for MCP agents that provides 25 tools for storing, searching, and managing agentic context. It uses hybrid search (combining semantic and keyword matching) and runs entirely on your machine with no cloud dependency.
Is my data private with m3-memory?
Yes. m3-memory is local-first and GDPR-compliant, meaning all memories stay on your device and you control data deletion. No data is sent to external servers.
How do I install m3-memory?
Install via uvx with the command: uvx m3-memory@2026.4.7. It runs over stdio transport, so it integrates directly with MCP-compatible clients.
What is hybrid search in m3-memory?
Hybrid search combines semantic (meaning-based) and keyword-based retrieval to find memories. This lets you search by concept or exact phrase, improving recall for agent context lookups.
Can I delete memories for GDPR compliance?
Yes. m3-memory includes GDPR-compliant deletion tools as part of its 25-tool suite, allowing you to purge specific memories or entire datasets on demand.
MCP Playground runs 10,000+ hosted MCP servers — GitHub, Linear, Notion, Stripe, Sentry and more — across Claude, GPT, Gemini, DeepSeek and 30+ AI models. Compare model answers side-by-side, save agent presets, share runs. Zero install.
Open Agent StudioJudges Panel
45 judges that evaluate AI-generated code for security, cost, and quality with built-in AST.
Addozhang Nexus
Query Sonatype Nexus Repository (OSS/Pro) for Maven, Python, and Docker artifacts
Cyanheads MCP Ts Template
TypeScript template for building MCP servers with declarative tooling, observability, and auth.
Chernistry Bernstein
Declarative agent orchestration for engineering teams