Monitors and optimizes the health of agent-based knowledge cycles, memory pools, and learning velocity.
The Flywheel Skill is a specialized operational tool for developers using AI coding agents. it monitors the 'knowledge flywheel'—the process by which session data is transformed into persistent, actionable knowledge. By measuring pool depths, extraction velocity, and artifact consistency, this skill ensures that your agentic workflows are compounding value over time rather than accumulating friction. It provides deep visibility into cache hit rates, detects stale information, and generates detailed health reports to maintain high-quality retrieval and agent performance.
主要功能
01Cache health analysis including hit rates and uncited learning detection
02332 GitHub stars
03Velocity tracking to monitor the rate of new knowledge extraction (last 7 days)
04Artifact consistency validation to identify broken internal references
05Automated knowledge pool measurement for learnings, patterns, and research
06Comprehensive health report generation with friction point identification
使用场景
01Auditing and pruning stale agent documentation to prevent knowledge drift
02Optimizing agent memory performance and retrieval quality in long-term projects
03Identifying bottlenecks in the automated learning-to-production knowledge pipeline
