From Forgetful to Brilliant.
Watch how AutoMem transforms your AI from a blank slate into a context-aware assistant.
Universal I/O
Compatible with Claude Code, Cursor, ChatGPT, OpenClaw, and any MCP client. Local or remote — works everywhere via MCP and Streamable HTTP.
Graph Engine
FalkorDB + Qdrant hybrid core. Maps relationships between memories, not just vector similarity.
Instant Recall
Sub-50ms latency. Your agent knows what you know before you finish typing the prompt.
It just clicks.
AutoMem runs quietly in the background, organizing your thoughts into a retrieval graph.
Dreams while you dream.
AutoMem runs quietly in the background, stitching every thought into a retrieval graph. It notices patterns, links context, and keeps the right edges warm so your agents answer with your voice—without you ever thinking about it.
0.11.0 • Embedding Providers & Memory Lookup
Voyage AI embedding provider for high-accuracy recall. OpenAI-compatible providers via OPENAI_BASE_URL (LM Studio, Ollama, xAI, etc.). GET /memory/<id> endpoint for direct memory lookup. Hardened time parser for numeric timestamps.
0.10.1 • Stability Fixes
Temporal query bound to 7-day window with timeout to prevent runaway queries. Skip temperature param for o-series and gpt-5 models.
0.10.0 • Visualization & Observability
Interactive graph visualization API. automem-watch real-time observability with SSE streaming and TUI. Ollama embedding provider for fully local deployments. Memory content size governance with auto-summarization. Streamable HTTP transport for MCP connections.
MCP 0.12.0 • OpenClaw Integration
Native OpenClaw integration replacing mcporter dependency chain. Direct HTTP API calls to AutoMem with new SKILL.md template. Simplified setup and troubleshooting documentation.
MCP 0.11.0 • Claude Code Plugin
Claude Code plugin with automatic memory capture hooks. Session start recall, session end summary. Cursor hooks integration for automatic capture. Integration tests and CLI smoke tests.
MCP 0.10.0 • MCP Registry & SessionStart Hook
Listed on the official MCP Registry. SessionStart hook for automatic memory recall at conversation start. Memory content size guidelines and warnings surfaced in the client.
0.9.3 • Expansion Filtering
expand_min_strength and expand_min_importance parameters for fine-grained graph traversal control. Vector dimension autodetect for embedding providers. CI workflow for automated testing on PRs.
MCP 0.8.1 • Spec Compliance
All MCP tools now return structuredContent with outputSchema, aligning recall outputs (results, dedup_removed) to spec.
MCP 0.8.0 • Advanced Recall Tools
Exposed expand_entities/relations, auto_decompose, context-aware boosts, and full tool metadata in the MCP client; simplified Claude Code integration.
0.9.1 • Entity Expansion + SOTA
Added expand_entities/entity_expansion params for graph hops and hit 90.53% on LoCoMo-10 (multi-hop up to 50%).
0.9.0 • Retrieval Engine Upgrade
Multi-hop bridge discovery, temporal alignment scoring, and weighted hybrid scoring (vector, keyword, relation, temporal, importance) boost recall precision.
0.8.0 • API Modularization & Security
Refactored monolithic app into modular blueprints, hardened memory IDs to server-generated UUIDs, and fixed Railway secrets/volumes for reliable deploys.
Open Source & Portable
Your memory infrastructure should be as portable as your code. Run it on your laptop, your private cloud, or our managed service.
Docker
The standard container. Run it locally or on any VPS.
automem/server:latest
Railway
One-click production deployment. Handles DBs & updates automatically.
> Success (24s)
Source
Fork the repo. Hack the graph logic. It's your code.
npm install