Skip to content

GBrain Memory Setup

Gives an existing AI agent a keyless, persistent memory store and walks through safe install or repair

Repository
garrytan/gbrain
Stars
31,447
Licence
MIT
Files
1 file

What it does

This skill guides an AI agent through installing GBrain, an open-source memory system, into an agent the user already runs. It starts with local keyless storage and keyword search, and explicitly avoids requiring API keys, a private repository, or background services. It asks the user to choose a search mode from a printed cost matrix, saves a test memory with a readback check, and reports separately whether the software installed, whether the agent's instructions were activated, and whether memory works across conversations.

Great for

  • Developers who want a personal memory layer for Claude Code, Codex or a similar coding agent
  • Users of Grok Bot or Muse who want their agent to remember things across chats
  • People who already have a hosted GBrain and need to connect a new agent to it

Skip it if

  • Users who expect automatic capture of all conversations; capture, imports and schedules need explicit opt-in and are not enabled by this setup
  • Anyone who wants a turnkey, zero-technical-knowledge install, since it runs shell commands and asks for choices about routing and databases
  • Environments where an existing database could be overwritten; the skill warns against reinitializing memory but mistakes in routing could still affect existing data

Try asking

  • “Set up GBrain memory for my Claude Code agent without any API keys”
  • “Add gbrain to my agent and keep my existing instructions and config untouched”
  • “Connect my agent to my existing hosted brain over MCP”
  • “Initialize a brain for this workspace using local PGlite storage”
Step-by-step guide: how to use GBrain Memory SetupInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the MIT licence and an attribution note.

    setup.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip setup.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip setup.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Review the printed search-mode cost matrix yourself before confirming; model API costs are separate from any harness subscription
  • Keep database credentials out of configuration files and let the agent use environment or private handoffs rather than copying secrets
  • Ask the agent to report software installed, instructions activated and cross-conversation recall as separate results so you can see what actually works

What's inside

  • SKILL.md

Redistributed unmodified from garrytan/gbrain under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.