gbrain
Garry's Opinionated OpenClaw/Hermes Agent Brain
GBrain is a persistent memory and knowledge-graph brain for AI agents that provides cross-session recall, synthesis, and automatic enrichment. Built by Y Combinator's president to power his personal agents managing 155,000+ pages, it combines hybrid search (vector + keyword + graph traversal) with a synthesis layer that produces cited answers rather than raw search results. It connects to Claude Code, Codex, OpenClaw, Hermes, ChatGPT, and other agent frameworks via MCP, offering both personal and team/company brain deployments with scoped access control.
Key Features
Use Cases
- 01Personal knowledge management with 150K+ page corpus for VC dealflow, people tracking, and meeting prep
- 02Company institutional memory with per-team scoped slices and federated source repositories
- 03Cross-session agent memory enabling persistent context across restarts and conversations
- 04Automated inbox processing identifying who's waiting on you and open commitments from email threads
- 05Voice-call brain integration via Twilio creating searchable conversation transcripts
- 06Procedural memory with Memorable integration capturing how tasks were solved for future replay
Related Agents
View morehermes-agent
The agent that grows with you
agency-agents
A complete AI agency at your fingertips - From frontend wizards to Reddit community ninjas, from whimsy injectors to reality checkers. Each agent is a specialized expert with personality, processes, and proven deliverables.
openinterpreter
A coding agent for open models like Kimi K3
cline
Autonomous coding agent as an SDK, IDE extension, or CLI assistant.
gbrain — FAQ
What is GBrain?+
GBrain is a production-grade AI agent memory system providing persistent knowledge storage, hybrid search, knowledge graph, and synthesis across sessions. It runs locally (PGLite) or on shared Postgres and connects to agent frameworks via MCP or CLI.
How do I install GBrain for my coding agent?+
Run 'bun install -g github:garrytan/gbrain', then 'gbrain init --pglite' for a local brain. For Claude Code: 'claude mcp add gbrain -- gbrain serve'. For Codex: 'codex plugin marketplace add garrytan/gbrain@codex-plugin' then 'codex plugin add gbrain@gbrain'.
Which AI clients and agents does GBrain work with?+
GBrain supports Claude Code, Codex, Cursor, Windsurf, Claude Desktop, ChatGPT, OpenClaw, Hermes, Grok Bot/Build, Muse, Perplexity Computer, and any MCP-compatible client via stdio or HTTP MCP server modes.
Do I need API keys to use GBrain?+
No, GBrain starts keyless with keyword search and agent-written memory. Optional API keys unlock semantic search (Voyage/OpenAI embeddings) and automatic fact extraction (OpenAI/Anthropic), but the core memory verbs work with zero external dependencies.
Is GBrain free and open source?+
Yes, GBrain is MIT-licensed open source. Embedding/LLM API costs are separate and optional; PGLite mode runs entirely offline with no server required.
How does the synthesis layer differ from regular search?+
Regular search returns ranked pages; GBrain's 'think' command reads those pages and produces a synthesized answer with citations and explicit gap analysis about what the brain doesn't know yet, eliminating the manual skim step.
How do I install gbrain?+
Open the source repository on GitHub and follow its README. gbrain is a agent — MCP Agents Market links you directly to the official repo.
Is gbrain free?+
gbrain is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.