I made a /wtf skill for Claude Code for when the agent leaves you saying "wtf did you just do?"
A user has developed a new tool called "/wtf" for Claude Code, designed to provide post-mortem analysis of agent actions. This tool aims to clarify what changes an AI made, why it made them, and what might have broken in the process. It offers a concise verdict, details on modifications, identifies missed checks, and provides rollback recipes and suggestions for future prompts. AI
IMPACT Provides a post-mortem analysis tool for AI code generation, helping users understand and correct AI actions.