What problem it solves

AI agents need a durable memory that records facts with provenance, supports corrections and withdrawal, and remains available across different agents. Conventional retrieval may surface relevant pages without synthesizing them into a sourced answer or identifying gaps in the available knowledge.

Key features

• Portable agent memory — Store explicit facts with sources and make them available across agents while keeping durable preferences and facts separate from transient harness configuration.

  Source: README

• Sourced answers and gap analysis — Ask for a synthesized answer across retrieved material, with citations and an explicit account of missing, stale, uncited, or contradictory information.

  Source: README

• Flexible knowledge schemas — Use bundled schema packs or derive custom types from an existing filesystem instead of requiring one fixed notes-and-people layout.

  Source: README

• Retrieval and graph linking — Start with keyless keyword retrieval and add semantic capabilities as needed. A typed knowledge graph can extract supported references from trusted local writes for relationship-aware navigation.

  Source: README

Quick start

Copied verbatim from the project's README.

bun install -g github:garrytan/gbrain#stable
gbrain setup claude-code              # or: gbrain setup codex
gbrain init --pglite --no-embedding   # any other harness, or no agent

gbrain init --pglite --no-embedding   # keyless; no server, no Docker
gbrain import ~/notes/ --no-embed     # index your markdown
gbrain search "a phrase from a note" --json
gbrain doctor                         # verify health

# raw retrieval: top pages by hybrid score, no answer-generation call
gbrain search "who's working on AI agents at portfolio companies?"

# brain layer: synthesized answer with citations and gap analysis
gbrain think "who's working on AI agents at portfolio companies?"

Who it may suit

Inference: may suit people who want to add a self-hosted, source-aware memory layer to an existing coding or personal agent. It may also suit teams evaluating a shared institutional knowledge system with explicit access controls.

Topics to discuss

• Which knowledge sources and correction workflows should the community prioritize for a shared agent memory?

• How should schemas balance automatic type detection with explicit human review?

• Which access and sharing scenarios should receive more guidance and verification examples?

Written by ReadmeX AI from the project's README, source and public GitHub data; it may contain mistakes.

Read the original README https://github.com/garrytan/gbrain#readme

ReadmeX opened this community for a trending project; it is not run by the project. Maintainers can claim it from the community page.