What problem does it solve? Coordinating periodic self-improvement across multiple IL agents (Owner, Researcher, Codifier, Librarian) is manual and inconsistent—reflections go stale, proposals lack traceability, and no record ties a round together. This Skill runs the full cycle: freshness checks, reflection triggering, focus-area dispatch, and proposal collection. ## Core Features & Use Cases - Reflection freshness checks: Scans each agent's agents/{name}/reflections/ directory and flags stale agents based on a 21-day / 3-session threshold or focus-area drift. - Shared reflection prompt injection: Loads an immutable, agent-agnostic prompt template and injects focus areas, agent name, and dates at run time. - Proposal collection with traceability: Each proposal in governance/proposals/ links back to its source reflection via derives_from_reflection, and a round record indexes the entire cycle. - Use Case: After a major system migration, Nick invokes a reflection round with focus areas like "context-rot" and "skill coverage"; each agent self-reflects, drafts proposals, and Nick reviews them before any become IB items. ## Quick Start Run a reflection round across all four IL agents with focus areas on efficiency and skill gaps, then summarize the proposals drafted for my review.
