Analyzes conversation history to automatically refine CLAUDE.md memory files and improve future AI interactions.
The Reflect skill serves as a self-improvement loop for your AI development workflow by analyzing recent conversation history for friction points, repeated corrections, and missing context. It identifies exactly where Claude's understanding diverged from your expectations and proposes surgical updates to your CLAUDE.md memory files—whether project-wide, personal, or local. By codifying these insights into persistent rules, it ensures that Claude learns from every interaction, reduces recurring mistakes, and maintains a high-fidelity understanding of your specific coding standards and project nuances.
주요 기능
01Built-in anti-pattern detection to keep memory files concise and effective
02Support for modular rule sets and hierarchical memory overrides
03Automated conversation analysis to identify friction and misunderstandings
0431 GitHub stars
05Interactive proposal system with checkboxes for selective rule updates
06Multi-tier memory management across project, user, and local files
사용 사례
01Capturing project-specific tribal knowledge and coding standards during development
02Permanently fixing repeated AI mistakes by codifying corrections into project rules
03Maintaining personalized global preferences that follow you across different repositories
