What problem does it solve? Autonomous agents accumulate stale memory, silent failures, and declining output quality over time without a structured feedback loop. This Skill performs a weekly self-audit so problems are caught and corrected instead of compounding. ## Core Features & Use Cases - Output Quality Audit: Reviews recent articles, notifications, and PR comments to judge whether outputs were substantive, useful, and actionable. - Reliability & Memory Hygiene Checks: Counts skills that ran versus expected, detects repeated error patterns, and verifies MEMORY.md stays current and under 50 lines. - Actionable Improvements: Generates recommendations for skills, schedules, and config changes, applies safe fixes directly (pruning stale entries, removing dead feeds), and sends a summary via the notify command. - Use Case: Run it every Sunday to review the past week's logs, prune outdated memory entries, and receive a notification summarizing reliability and the top improvement suggestions. ## Quick Start Run the self review skill to audit this week's agent activity and send me a summary of quality, reliability, and recommended improvements.
