rule-auditlisted

Find out which of your agent rules are actually enforced and which are only hope — then convert the ones that matter into deterministic enforcement. Use whenever the user says Claude ignored my rules, it pushed to main again, it keeps forgetting my instructions, my CLAUDE.md isn't working, "same mistake again", "I told it not to", "rule follow nahi karta", "phir wahi ghalti" — or asks to add a hook, block a command, enforce a convention, run something automatically after edits, review their CLAUDE.md, or make an instruction actually stick. Also use before writing a new rule, to decide whether it belongs in prose or in the harness, and when a CLAUDE.md has grown long enough that compliance is degrading. It inventories every rule you have, classifies each as enforceable or advisory, generates the hook or permission config for the enforceable ones, tells you plainly which rules can never be enforced, and prunes the rest so the remaining instructions actually get followed.
xniperbuilds/rule-audit · ★ 0 · Code & Development · score 72

Install: claude install-skill xniperbuilds/rule-audit

# Rule Audit > **An instruction is a request. A hook is a guarantee.** > Rules that must never break do not belong in prose. They belong in the harness. ## The problem this exists for Writing a rule in CLAUDE.md feels like fixing the problem. It isn't — it's *asking* for the problem to be fixed, on every future turn, forever, under context pressure. What is actually measured: - CLAUDE.md rules are followed roughly **70% of the time** — meaning a "never push to main" rule is a production incident on a schedule. - **Instruction compliance decreases as instruction count increases.** Adding a rule to fix a violation makes every *other* rule slightly less likely to be followed. This is the trap: the natural response to a broken rule is to write another rule, which makes it worse. - Long sessions, ambiguity, and content pulled in from files all push compliance down further. - Users have asked Anthropic directly for this (`anthropics/claude-code` issue **#32163** — "Hard-enforce CLAUDE.md rules via code — rules marked CRITICAL should trigger hooks, not just prompts"). Hooks exist; the *conversion of your prose rules into them* does not. *(Figures are from secondary sources; the mechanism facts in `references/hook-reference.md` were verified against the official reference and a working config on 2026-07-26. See "Sourcing" at the end.)* So the core move is not "write better rules". It is: **decide which rules must be mechanical, make those mechanical, and delete enough of the r