mcp-memory-service
Open-source persistent memory for AI agent pipelines (LangGraph, CrewAI, AutoGen) and Claude. REST API + knowledge graph + autonomous consolidation.
What is mcp-memory-service?
Open-source persistent memory for AI agent pipelines (LangGraph, CrewAI, AutoGen) and Claude. REST API + knowledge graph + autonomous consolidation.
Overview of mcp-memory-service
What it is
Open-source memory backend for AI agents — REST API, MCP, OAuth, CLI, dashboard. One self-hosted service, every transport. Agents store decisions, share causal knowledge graphs, and retrieve context in 5ms — without cloud lock-in or API costs. Works with LangGraph, CrewAI, AutoGen, any HTTP client, Claude Desktop, OpenCode.
Key features
- Framework-agnostic REST API — no MCP client library needed; the live endpoint list is at
/api/docs - Knowledge graph — agents share causal chains, not just facts
X-Agent-IDheader — auto-tag memories by agent identity for scoped retrievalconversation_id— bypass deduplication for incremental conversation storage- SSE events — real-time notifications when any agent stores or deletes a memory
- Embeddings run locally via ONNX — memory never leaves your infrastructure
- Autonomous consolidation — schedule clustering, compression, association discovery, and controlled forgetting
Use cases
Persistent memory for AI agents across sessions; sharing context among multiple agents in pipelines; capturing project decisions and code patterns automatically; avoiding re-explaining context in new chats. Works with agent frameworks (LangGraph, CrewAI, AutoGen), CLI and terminal tools (Claude Code, Gemini CLI, OpenCode), desktop and IDE (Claude Desktop, VS Code, Cursor, Windsurf, Raycast, JetBrains, Zed), and browser chat (claude.ai, ChatGPT via Remote MCP over HTTPS).
Getting started
pip install "mcp-memory-service[sqlite]"
memory server
For HTTP API mode (agent pipelines):
MCP_ALLOW_ANONYMOUS_ACCESS=true memory server --http
License and pricing
License: Apache-2.0. Open source (free).
Based on the project's public README and repository metadata.