| name | setup-verify |
| description | Runs the 10-point environment health check the playbook uses to gate White Belt's Quest W-0. Produces GREEN / YELLOW / RED status per check with one-line fixes for any non-green result. Activate when the user runs the setup script, finishes onboarding, or asks "am I set up correctly". |
setup-verify
Triggers
Activate when the user says "run setup-verify", "verify my setup", "am I set up correctly", or "what colour am I". Activate proactively when the user has just finished running the setup script. Activate when the user reports that something is not working and the cause might be environmental.
Body
The skill runs ten checks in order. Each check produces one of three statuses.
- GREEN. The check passes. No action needed.
- YELLOW. The check passes but with a caveat (version slightly behind pinned, optional component missing, non-blocking warning). The setup is functional; the user should fix when convenient.
- RED. The check fails. The user cannot proceed with the curriculum until this is resolved.
The ten checks are documented in checks.md. Each check has a fix documented in one-line-fixes.md. The output format is documented in output-shape.md.
Workflow
- Read the user's environment state. Use the shell to query versions, run health probes, and check installed components. Do not modify the environment.
- Run each of the ten checks in order. Record status (GREEN / YELLOW / RED) and the specific signal (version number, error message, command output) that produced it.
- For each non-green check, look up the one-line fix from
one-line-fixes.md.
- Produce a structured report per
output-shape.md. The report is ready to paste into #ai-help if the user is YELLOW or RED after one hour of attempting fixes.
