gbrain

Memory Stargraph is a local web service for visualizing and exploring a GBrain knowledge base as an interactive star-cloud entity graph, enabling AI agents to share and manage memory across hosts.

Garry Tan shares the complete architecture of GBrain, an AI model.

Garry Tan introduces GBrain, a system designed to manage large-scale context for AI agents by loading the most relevant information into their limited context window, likening it to a 'Postgres for agents'.

GBrain v0.42.30 now provides a detailed summary of how your thinking has changed over time.

Shann Holmberg describes a structured approach to building an AI agent company within an agency, using a central brain (gBrain), an orchestrator agent (Hermes), and narrow-scoped specialist agents for different departments, with isolated client pods to prevent context leakage.

Garry Tan's gbrain-evals is an open-source test suite for gbrain, an AI agent's long-term memory, with 4 end-to-end evaluations verifying SkillOpt functionality, achieving high recall and precision on multiple benchmarks.

This thread argues that AI's fundamental problem is memory, not intelligence, and highlights GBrain as a system designed to provide persistent, synthesizing memory for AI agents, featuring uncertainty surfacing and biological-like memory consolidation during sleep.

Garry Tan shares ongoing improvements to gbrain, a retrieval tool for personal and company brains, noting weekly advancements.

The latest version of GBrain adds synthesized answers to specific queries, moving beyond basic retrieval, as shown in an AB test.

GBrain v0.40.0 introduces a voice agent based on Gemini Live for OpenClaw/Hermes Agent + GBrain, enabling voice capabilities with large context and tool use.

A quote tweet discusses gBrain being state-of-the-art for a specific use case, with a shared memory layer architecture under Hermes Agent.

Garry Tan shares his favorite uses of GBrain, including personalized book rewriting and a brainstorming function called LSD (lateral synaptic drift) that mashes ideas together.

Shann Holmberg describes an experimental architecture using gBrain as a shared memory layer for a team of Hermes Agents, allowing specialists to read from a centralized brain before acting and write durable context back.

Discusses how to avoid cold starts for the Hermes AI assistant through explicit accumulation (AGENTS.md, Skill) and implicit accumulation (memory, session search), so it truly becomes a personal system, citing GBrain as supporting evidence for a personal knowledge base.

Garry Tan promotes the concept of process power as a unique moat enabled by personal AI and his GBrain project, claiming anyone can build their own mini-AGI.

GBrain can now package skills, code, tests, and evals into a SKILLPACK tarball that others can install using a simple command.

Garry Tan explains the motivation behind releasing his X articles on AI agents and concepts like GStack and GBrain, emphasizing the idea of 'process power' as a personal moat.

GBrain v0.36.1 adds full evals inspired by a hackathon team's Hindsight concept for tracking predictions over time.

Garry Tan demonstrates a metaprompting technique to rewrite a README for GBrain by evaluating it from a novice user's perspective and iteratively improving it to a perfect score.

GBrain v0.36 is released with a new README and a revamped skillpack system that allows users to customize and update skill bundles independently.