Labs / workspace-memory
Workspace Memory
workspace-memory is an open source memory and governance layer for AI-assisted software projects. Human contract files remain the reviewed truth, Neo4j stores primary graph memory and temporal reasoning state, and PostgreSQL records the operational ledger. Agents get focused, cited, graph-first context without guessing from chat history or treating generated memory as human-approved truth.
Contribution
I designed and built workspace-memory as a local-first system for keeping software intent, implementation evidence, temporal graph memory, and agent context aligned. The project uses strict Neo4j/PostgreSQL role separation: Neo4j owns knowledge, relationships, lifecycle history, search, context, impact, and code evidence; PostgreSQL stores source events, document versions, audit records, graph write receipts, and operational
Programming languages
Python
Frameworks and tools
FastAPI, Neo4j, OpenAI API, Pydantic, Uvicorn