What problem does it solve? Knowledge-brain deployments drift over time: versions go stale, schema migrations pile up, jobs stall, and embedding coverage drops. This Skill surfaces those issues proactively as a ranked, severity-tagged list so the owner knows exactly what to fix and how, without digging through logs. ## Core Features & Use Cases - Ranked health findings: Runs gbrain advisor --json and returns findings with severity (critical/warn/info), a one-line explanation, and the exact fix command for each issue. - Read-only by contract: Never mutates brain state; every fix is shown to the user and executed only after explicit approval, optionally via gbrain advisor --apply <dispatch_id>. - Weekly cadence checkup: Includes a cron recipe for a weekly brain checkup that reports only what is new since the last run, avoiding repeated nagging. - Use Case: On Monday morning, the scheduled job detects pending schema migrations and an available gbrain upgrade, then pings the user with the two findings and their exact fix commands, asking before running anything. ## Quick Start Ask the assistant to run a gbrain advisor checkup and summarize the top findings with their fix commands.
