What problem does it solve? Once the pi-events-daemon has been running in real usage, known risks (parked backlog, dead-letter growth, wake failures, low adoption) can silently degrade without any telemetry to surface them. This Skill carries a persistent risk register so discovered risks are never re-derived from scratch, and runs evidence-based periodic diagnostics on the daemon. ## Core Features & Use Cases - Risk Register Maintenance: Tracks 10+ known daemon risks (R1-R10) plus environmental migration risks in references/risks.md, updating status and dates after every diagnosis. - Evidence-Based Diagnosis: Runs read-only SQLite queries, journalctl log audits, and transcript sampling per references/checks.md, issuing verdicts of healthy, degraded, or triggered only with supporting evidence. - Remediation Routing: Applies immediate fixes where safe (FTS5 index rebuild, dead-letter cleanup) and escalates structural issues to the issue tracker with links recorded in the register. - Use Case: After the daemon has run for a few weeks, a user notices events are not being delivered. Invoke this Skill to check parked delivery ages, dead-letter reasons, and wake accepted-to-verified latency, then record findings back into the risk register. ## Quick Start Run a health diagnosis on the pi-events-daemon and update the risk register with verdicts for each known risk.
