Feedback Loop - Self-Improvement

To capture runtime signals for self-improvement, run the post-task review workflow after finishing a feature. This logs structured learnings and suggested improvements to skill docs and patterns for actionable updates.

Mid-task gap detection identifies missing patterns or outdated guidance during active execution. It captures these runtime signals to trigger immediate improvements across skills within the AI library.

Post-task reviews generate structured learnings by analyzing completed work. They produce suggested improvements to skill documentation and patterns, creating standardized entries for library maintenance.

Yes, periodic maintainer signals aggregate captured runtime insights to enable library evolution and quality checks. They compile mid-task gaps and post-task learnings into actionable maintenance updates.

Yes, standardized review workflows are required to generate actionable updates. The self-improvement process depends on clear signal types and structured improvement entries to trigger effective maintenance.