agent-analytics
How often do you correct your coding agent, and on what kind of work? Measures it from your Claude Code transcripts, drafts CLAUDE.md rules from your recurring corrections, and exports them as preference pairs.
