A self-hosted temporal memory platform for AI agents and apps
TMCRA provides a structured memory architecture that separates durable evidence, fast recall, and slow semantic consolidation. It prevents AI hallucinations by returning source-aware context rather than raw text. Designed for developers building multi-tenant agentic systems, it offers a full stack including an API, SDKs, and an MCP server.
View on GitHub →reshuibuduo/tmcra