Skip to content

Smoke Test Health Checker

Checks gbrain install health after restarts and auto-fixes bounded local issues

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

What it does

When installed, the agent can run a set of health checks on a gbrain setup (and OpenClaw services if present), typically after a computer or container restart or when something seems broken. It tests 8 core items such as the Bun runtime, the gbrain CLI and database, the worker process, the embedding API key and the brain repo, tries simple local fixes for some failures, and reports what passed, failed or was skipped. It can also run your own check scripts placed in a user folder.

Great for

  • Developers who run gbrain on a local machine or container and restart it often
  • Operators of OpenClaw deployments who want a startup check for services
  • Anyone who wants a quick diagnosis after something stops working

Skip it if

  • People who do not use gbrain, since most checks are specific to it (not useful and may report misleading skips)
  • Situations where you do not want automatic changes made to packages or the database, because the skill is marked as mutating and runs a database repair tier when the user agrees (risky on production data without a backup)

Try asking

  • “Run a smoke test, my container just restarted”
  • “Health check gbrain and tell me what failed”
  • “Did the restart break anything?”
  • “Run smoke tests and show me the unfixed failures”
Step-by-step guide: how to use Smoke Test Health CheckerInstall, 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.

    smoke-test.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

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

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip smoke-test.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

  • Run it once per restart or on request rather than on every chat turn, since it is cheap but not free.
  • Keep custom drop-in scripts in ~/.gbrain/smoke-tests.d/ and wrap every command in timeout so one hung script does not block the rest.
  • Read the report carefully: skipped checks are not passes, and the exit code counts only real unfixed failures.

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.