Open Brain
A user-owned, model-independent memory layer for coding agents that stores durable evidence, decisions and outcomes in PostgreSQL with pgvector and assembles compact context for the next task.
Open Brain gives coding agents a durable memory layer that is independent of any one model. It stores semantic memory in PostgreSQL and pgvector, preserves where each piece of knowledge came from, follows work across sessions and agents, and assembles compact context for the next task. It is the memory and intelligence layer, not the agent runtime; a connected agent still handles reasoning and execution itself.
It separates canonical evidence such as append-only events, sessions and tool results, a knowledge model of current assertions, projects, tasks and decisions, and rebuildable retrieval projections such as embeddings and caches. Imported or provider-supplied records are never silently promoted to fact; they keep their provenance until a reconciliation step classifies them. It has native integrations for Hermes, Medusa, Codex and Claude Code, and connects to other agents through REST, MCP or a CLI.