A developer measured what my Claude.md, skills and hooks are worth

A developer measured what my Claude.md: what actually changed

HN Claude/Codex/Fable reports: I measured what my Claude.md, skills and hooks are worth. Article URL: https://github.com/jameskomo/config-drift-checker Comments URL: https://news.ycombinator.com/item?id=49468945 Points: 2 # Comments: 0

A developer measured what my Claude.md: how it works

This turns those rules into eval cases, runs them on every PR and every release against a pinned baseline, and tells you the moment something stops working: when, why, and what moved. First-class on Claude Code (skills, hooks, release canaries, the whole drift machinery); the same cases also run through OpenAI's Codex and Google's Gemini CLIs, with live passing runs on all three, experimental labels on the newer two until full published comparisons.

A developer measured what my Claude.md: why it matters now

Proof it works: we broke our own setup, and the repair skill fixed it. A skill's trigger description rewritten the way a careless PR would: the suite fell 1.00 → 0.56, the tripwire case read 0.00, and the report named the cause itself: the skill was discovered but never invoked, so fix the trigger wording, not the packaging.

A developer measured what my Claude.md: who is affected

Then the repair skill restored the behaviour on its own and proved it with a green re-run, for $0.28. See the full write-up from HN Claude/Codex/Fable via the source link for quotes and complete context.

A developer measured what my Claude.md: what to watch

Each is a single file with no dependencies (read it first, it's short): curl -fsSLO https://raw.githubusercontent.com/jameskomo/config-drift-checker/v1/config-drift-checker/tools/skill-lint.mjs && node skill-lint.mjs . # your SKILL.md files curl -fsSLO https://raw.githubusercontent.com/jameskomo/config-drift-checker/v1/config-drift-checker/tools/suite-doctor.mjs && node suite-doctor.mjs .