What problem does it solve? Operators of a gbrain install lack a single agent-readable signal for whether the skillpack is healthy, forcing them to manually run doctor checks and migration listings and interpret the results themselves. ## Core Features & Use Cases - Unified health report: Wraps gbrain doctor and gbrain apply-migrations --list into one JSON payload with healthy, summary, actions, doctor, and migrations fields. - Exit-code gating: Returns 0 when healthy, 1 when remediation is needed, and 2 when the state cannot be determined, so crons and CI pipelines can gate on the code alone. - Actionable remediation: The actions[] array lists the exact commands to run (e.g. gbrain apply-migrations --yes, gbrain embed --stale, gbrain check-backlinks fix). - Use Case: Wire gbrain skillpack-check --quiet into a morning-briefing cron so the agent surfaces a one-line alert only when the brain needs attention. ## Quick Start Ask your agent to run gbrain skillpack-check and report whether the install is healthy along with any remediation commands.
