Persistent memory for AI agents. Semantic search, knowledge graph, automatic consolidation — now accessible from Claude.ai via Streamable HTTP.
OAuth 2.1 refresh_token grant with atomic rotation (MCP SEP-2207) — plus knowledge graph on streamable-http. 1,692 tests.
Clients requesting offline_access scope receive a refresh token with atomic rotation. Every refresh issues a new token and revokes the old one. Replay detection walks the full token chain and bulk-revokes descendants. (PR #766, @netizen1119)
The memory_graph tool (connected memories, shortest path, subgraph extraction) now works in the FastMCP streamable-http transport, matching stdio parity. Shared GraphService layer for both server variants. (PR #759, @henry201605)
Defense-in-depth guards against CodeQL py/reflective-xss and py/url-redirection in the authorization-code redirect flow. Dangerous schemes rejected; meta-refresh and JS redirect strings properly escaped. (PR #745)
Claude.ai connects to your memory server through Streamable HTTP with OAuth 2.1 authentication.
Battle-tested with comprehensive testing and optimized for performance.
Install from PyPI and connect your AI agents to persistent memory.