
Project Test Suite Health Checker
Validates agent skill conformance and runs the project test suite, classifying failures as regressions, stale tests, flakes or infra issues.
- Repository
- garrytan/gbrain
- Stars
- 31,447
- Licence
- MIT
- Files
- 1 file
What it does
This skill makes an agent act as a quality checker for the gbrain project. In its first mode it checks that every skill folder has a correctly formatted SKILL.md, that manifest and resolver files list every skill, and that no two skills claim the same trigger phrases. In its second mode it runs the project's tiered tests (unit, evals, integration, system health), reads the git history to work out why any test failed, sorts each failure into a category, and writes a daily report. It also keeps a history file to track trends over time.
Great for
- Maintainers of the gbrain repository who want a daily summary of test health
- Developers who want the agent to diagnose why tests broke after a code change
- Contributors adding new skills who need to check they conform to the project's structure
Skip it if
- Users who just want to use gbrain for everyday tasks, since this only serves project maintenance
- Projects other than gbrain, because the paths, manifests and commands are specific to this repository
- Unattended auto-fixing of tests, because the skill itself warns that changing a test to match changed behaviour can hide a real bug
Try asking
- “Validate skills”
- “Run the tests and tell me what's broken”
- “How are the tests doing today?”
- “Run a skill health check”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 1 original files, the MIT licence and an attribution note.
testing.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip testing.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip testing.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 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
- Ask for a single mode explicitly (for example 'validate skills' or 'run the tests') so the agent picks the right workflow.
- Review any proposed test edits before accepting them, especially where the agent suggests the code may be wrong instead.
- Expect the daily run to write results to ~/.gbrain/test-state.json, so check that location if you want a history.
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.