GenAI PMGBrain
A GitHub repository shared by Garry Tan that packages skills and a knowledge-wiki style setup. Relevant to AI PMs interested in personal knowledge systems and reusable skill repositories.
Key Highlights
- GBrain is positioned as an open-source retrieval and memory layer for AI agents built around markdown knowledge bases and reusable skills.
- Its evolution spans retrieval performance claims, MCP compatibility, embedding stack changes, and support for Codex and Claude Code.
- For AI PMs, GBrain offers a practical pattern for internal copilots, team knowledge systems, and repeatable agent-assisted workflows.
- A major framing of the project is the combination of a personal or company brain with a Karpathy-style knowledge wiki.
- The GBrain ecosystem includes related evaluation artifacts, agent frameworks, and repository-based operating models for AI-enabled teams.
GBrain
Overview
GBrain is an open-source retrieval, memory, and knowledge-wiki tool shared by Garry Tan, designed to help AI agents and humans work against large collections of markdown-based knowledge. Across newsletter mentions, it is described both as a retrieval engine for agentic systems and as a practical repository structure that packages reusable skills plus the beginnings of a “Karpathy-style knowledge wiki.” It has been positioned as MIT-licensed, GitHub-native, and compatible with agent frameworks such as OpenClaw, Hermes Agent, Codex, and Claude Code.For AI Product Managers, GBrain matters because it sits at the intersection of personal knowledge management, agent memory, enterprise retrieval, and workflow standardization. It is relevant not just as a standalone tool, but as a pattern: structuring institutional knowledge into reusable, searchable assets that can be accessed by AI systems without heavy query rewriting. That makes it useful for PMs thinking about internal copilots, company-brain architectures, evaluation of retrieval quality, and durable knowledge systems that scale with teams.
Key Developments
- 2026-05-18: GBrain began recommending ZeroEntropy as its default embedding and re-ranking engine, replacing OpenAI and Voyage AI.
- 2026-05-20: Garry Tan released GBrain as an MIT-licensed open-source retrieval and memory system, claiming state-of-the-art performance on LongMemEval without LLM query rewriting.
- 2026-05-24: GBrain was described as a state-of-the-art retrieval engine for agents, built for OpenClaw and Hermes with MCP server support for broader agent-harness compatibility. An update also added synthesized answers alongside basic retrieval, with references to improving accuracy in A/B tests between GBrain Search and GBrain Think.
- 2026-06-01: Garry Tan open-sourced GBrain on GitHub and outlined a roughly 30-minute setup based on his large markdown LLM wiki and an OpenClaw/Hermes agent that automates many tasks.
- 2026-06-06: GBrain was framed as a modular “company brain” framework that organizes work through scoped AI agents and client pods, suggesting a broader operating model beyond retrieval alone.
- 2026-06-16: Garry Tan introduced GBrain as an MIT-licensed retrieval layer for AI agents, likening it to a just-in-time “3-book librarian,” and noted adoption by thousands of developers.
- 2026-07-21: GBrain was launched as a free open-source retrieval library optimized for Hermes Agent and OpenClaw, with support for Codex and Claude Code, and described as powering Garry Tan’s personal company brain and AI.
- 2026-07-28: Garry Tan stated that GBrain had reached state-of-the-art agentic retrieval performance without LLM rewriting, with evaluation artifacts published in the gbrain-evals GitHub repository.
- 2026-08-13: GBrain v0.45.6.0 was released with 17 new brain skills, reportedly hardened through use with a personal OpenClaw agent over hundreds of thousands of markdown files. Compatibility with Codex and Claude Code was highlighted.
- 2026-08-18: Garry Tan shared the GBrain GitHub repository as a private-repo package containing 70 “proven skills” plus the beginnings of a “Karpathy-style knowledge wiki,” emphasizing that it is free, open source, and MIT-licensed.
Relevance to AI PMs
1. Prototype internal knowledge copilots faster: GBrain offers a concrete blueprint for turning scattered markdown docs, notes, and workflows into a searchable memory layer for AI agents. PMs can use this pattern to pilot team assistants for product specs, customer research, incident notes, and roadmap context.2. Design better retrieval systems without over-relying on prompt tricks: A recurring theme in GBrain’s positioning is strong retrieval without LLM query rewriting. For PMs evaluating RAG systems, this is tactically useful because it shifts focus toward corpus structure, embeddings, reranking, and evaluation instead of just prompt engineering.
3. Operationalize reusable product and ops skills: The “proven skills” framing is especially relevant for AI PMs building repeatable workflows. Teams can package recurring tasks—PRD drafting, launch prep, competitor analysis, support triage, postmortems—into reusable skill artifacts that are discoverable by both humans and agents.
Related
- garry-tan: Creator and primary public advocate of GBrain.
- github: GBrain is distributed as an open-source GitHub repository and presented as GitHub-native knowledge infrastructure.
- openclaw / hermes-agent / hermes: Core agent environments GBrain is repeatedly described as optimized for.
- claude-code / codex / opus: Developer-facing agent tools and models GBrain is described as supporting or integrating with.
- mcp: GBrain includes MCP server support, making it easier to connect to broader agent ecosystems.
- rag: GBrain is highly relevant as a retrieval and memory layer within RAG-style architectures.
- zeroentropy / openai / voyage-ai: Embedding and reranking providers connected to GBrain’s retrieval stack evolution, with ZeroEntropy later recommended by default.
- company-brain / client-pods: Concepts associated with the broader organizational architecture around GBrain, where scoped agents support structured workflows.
- gbrain-evals: Evaluation repository cited as evidence for retrieval performance claims.
- git-history / dry / gstack: Adjacent tooling and workflow concepts likely relevant to teams exploring repository-based knowledge systems and agent operations.
- karpathy-style-knowledge-wiki: A key framing for how GBrain structures durable, markdown-based institutional knowledge.
- hyojun_at: Related entity mentioned alongside the broader ecosystem around GBrain.
Newsletter Mentions (17)
“Garry Tan shared the gbrain GitHub repository, describing it as offering a private GitHub repo with 70 of his “proven skills” and the beginnings of a “Karpathy-style knowledge wiki.””
GenAI PM Daily August 18, 2026 GenAI PM Daily 🎧 Listen to this brief 3 min listen Today's top 20 insights for PM Builders, ranked by relevance from X, YouTube, LinkedIn, and Blogs. Cursor releases Origin, its integrated code hosting platform #1 𝕏 Cursor released Origin, its code hosting platform, with deep Cursor integration and repository syncing from GitHub. Cursor describes Origin as fast and easy to use. Also covered by: @Cursor , @Guillermo Rauch #2 𝕏 Philipp Schmid demonstrated Gemini 3.7 Flash using his Android emulator via ADB for the task “Play 1 round of Wordle.” He said its latency and visual reasoning make it exceptionally good for multimodal agentic use cases such as mobile control and Computer Use. #3 𝕏 LlamaIndex 🦙 recapped ExtractBench, which requires both extracted values and citations to be correct and evaluates word-level boxes at IoU 0.5. LlamaExtract Agentic Plus led with 84.9% page-level and 46.4% word-level results, achieving 87.1% on long documents where other systems scored zero. #4 𝕏 Garry Tan shared the gbrain GitHub repository, describing it as offering a private GitHub repo with 70 of his “proven skills” and the beginnings of a “Karpathy-style knowledge wiki.” The material is free, open source, and MIT-licensed, with full documentation in the README.
“Garry Tan released GBrain v0.45.6.0 with 17 new brain skills, hardened through his personal OpenClaw agent using hundreds of thousands of markdown files.”
#15 𝕏 Garry Tan released GBrain v0.45.6.0 with 17 new brain skills, hardened through his personal OpenClaw agent using hundreds of thousands of markdown files. GBrain now works with Codex and Claude Code.
“Garry Tan demonstrates that GBrain is now state-of-the-art for agentic retrieval without any LLM rewriting, with full evaluation results in the gbrain-evals GitHub repo.”
GenAI PM Daily July 28, 2026. GBrain is discussed as a retrieval breakthrough with supporting evaluation artifacts.
“Garry Tan launched GBrain, a free open-source retrieval library optimized for Hermes Agent and OpenClaw (with Codex and Claude Code support) that powers his personal company brain and AI.”
Presented as an open-source retrieval layer for agentic knowledge access and personal AI memory.
“Garry Tan introduced GBrain, an MIT-licensed open-source retrieval layer for AI agents—acting like a Postgres-style, just-in-time “3-book librarian”—and it’s already in use by thousands of developers.”
#9 𝕏 Garry Tan introduced GBrain, an MIT-licensed open-source retrieval layer for AI agents—acting like a Postgres-style, just-in-time “3-book librarian”—and it’s already in use by thousands of developers.
“Summary: Garry Tan unveiled GBrain, a modular “company brain” framework that structures work via scoped AI agents organized into client pods.”
#12 𝕏 Summary: Garry Tan unveiled GBrain, a modular “company brain” framework that structures work via scoped AI agents organized into client pods. This detailed agent company architecture standardizes workflows and scales cross-functional collaboration.
“Garry Tan open-sourced GBrain (MIT-licensed) on GitHub and outlines a 30-minute setup using his 350k-page markdown LLM wiki plus an OpenClaw/Hermes agent that automates most tasks.”
#6 𝕏 Garry Tan open-sourced GBrain (MIT-licensed) on GitHub and outlines a 30-minute setup using his 350k-page markdown LLM wiki plus an OpenClaw/Hermes agent that automates most tasks.
“Garry Tan launched GBrain, an MIT-licensed, state-of-the-art retrieval engine for agents—built for OpenClaw and Hermes but with full MCP server support to plug into almost any agent harness.”
#6 𝕏 Garry Tan launched GBrain, an MIT-licensed, state-of-the-art retrieval engine for agents—built for OpenClaw and Hermes but with full MCP server support to plug into almost any agent harness. #10 𝕏 Garry Tan rolled out the latest GBrain update, which adds synthesized answers to your queries instead of just basic retrieval. An A/B test of GBrain Search vs. GBrain Think shows it improving in accuracy every single day.
“Garry Tan released GBrain—an MIT-licensed OSS retrieval and memory system that achieves SOTA on LongMemEval, beating all known open-source repos by over 1% without any LLM query rewriting.”
#12 𝕏 Garry Tan released GBrain—an MIT-licensed OSS retrieval and memory system that achieves SOTA on LongMemEval, beating all known open-source repos by over 1% without any LLM query rewriting. He’s already running it on his own 100k-page OpenClaw/Hermes Agent brain.
“#6 𝕏 Garry Tan announces that GBrain now ships with ZeroEntropy as its recommended default embedding and re-ranking engine, replacing OpenAI and Voyage AI.”
#6 𝕏 Garry Tan announces that GBrain now ships with ZeroEntropy as its recommended default embedding and re-ranking engine, replacing OpenAI and Voyage AI.
Related
Anthropic’s coding agent used in a hardware/IoT experiment on a Raspberry Pi. It is relevant as an example of code-generation applied to physical systems and automation.
OpenAI is the company discussed here in relation to a research model security incident and mitigation changes. The newsletter frames the event as a warning shot about collaborative agent risks.
An AI coding agent or environment mentioned as a place to run AI eval skills. It is also listed as one of the agents that can be compared in a shared environment.
A standardized agent test suite referenced for model evaluation. The newsletter cites success rates on OpenClaw as part of the Nemotron benchmark result.
Founder and Y Combinator leader mentioned for reducing token load in GStack, highlighting cost/performance optimization for AI systems.
An interoperability protocol for connecting AI systems and tools. Here it is described through a public roadmap covering long-running workloads, local-server HTTP, discovery, identities, permissions, and generated SDKs.
An embeddable assistant capable of streaming answers, rendering UI, and acting within applications.
A software development platform used here as the source and sync target for repositories. It is central to AI coding workflows, plugin distribution, and agent automation.
RAG is a retrieval-based pattern that injects external context into prompts to improve model responses. The newsletter presents it as often outperforming fine-tuning for practical product work.
A model used in the newsletter as a reasoning and execution engine for product experimentation. It is described as generating daily A/B test ideas and implementing winners for a mobile game economy.
An AI agent environment or product that can host models and persona features. In this newsletter it appears both as a place where Qwen3.8-Max is available and as a tool with a /personality feature.
Stay updated on GBrain
Get curated AI PM insights delivered daily — covering this and 1,000+ other sources.
Subscribe Free