Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Rule Enforcement Map — Find Rules With No Hook Behind Them

    1

    A rule in `.claude/rules/` that nothing enforces is a suggestion.

    $15

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Rule Enforcement Map — Find Rules With No Hook Behind Them

    Rule Enforcement Map — Find Rules With No Hook Behind Them

    Example session with this skill installed

    Run Rule Enforcement Map on this project's .claude/rules folder (include the tests folder with --extra-dir tests) and tell me which rules nothing enforces, and which ones point at a script that does not exist.

    • Read your context and instructions
    • Compiled the rule enforcement map

    Summary

    5 rules found in .claude/rules/. 3 are enforced by something other than the agent's memory, 1 is honor-system only, and 1 claims an enforcer that does not exist.

    Result

    Rule Enforcement Map: 5 rules, 3 enforced, 1 honor-system, 1 broken-claim
    
    !!  commit-message-style  nothing mentions this rule
    BAD docs-stay-current     declared enforcer missing: scripts/check_docs.py
    OK  no-force-push         hook-script
    OK  no-secrets-in-logs    extra
    OK  tests-before-done     hook-script
    
    RuleStatusWhy
    no-force-pushOKa hook script mentions it
    tests-before-doneOKa hook script mentions it
    no-secrets-in-logsOKfound in an extra folder you named
    commit-message-stylehonor systemno hook, script, git hook, CI job or test mentions it
    docs-stay-currentbroken claimit declares enforced-by: scripts/check_docs.py, but that file is missing

    Next steps

    1. Create scripts/check_docs.py or remove the enforced-by comment from docs-stay-current.
    2. Decide whether commit-message-style should get a git hook; if not, leave it as a guideline.
    3. Add --fail-on-honor-only to CI to stop new unenforced rules from slipping in.

    Note: a match means something mentions the rule by name, not that the hook works.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    A rule in .claude/rules/ that nothing enforces is a suggestion. The model follows it when it remembers and ignores it when it does not, and nothing tells you which rules are which.

    Rule Enforcement Map answers the question in one command:

    Rule Enforcement Map: 5 rules, 3 enforced, 1 honor-system, 1 broken-claim
    
    !!  commit-message-style  nothing mentions this rule
    BAD docs-stay-current     declared enforcer missing: scripts/check_docs.py
    OK  no-force-push         hook-script
    OK  no-secrets-in-logs    extra
    OK  tests-before-done     hook-script
    

    What it does

    • Reads every rule file and searches the places that can enforce it: hook commands in settings.json and settings.local.json, the scripts those hooks run, .claude/hooks/, git hooks, CI workflows, and any folder you name with --extra-dir (tests, scripts)
    • Marks each rule OK, !! (honor system) or BAD (it declares <!-- enforced-by: path --> but the path is missing)
    • Lets you declare enforcers the name match cannot see: an enforced-by comment or enforce_keywords in the front matter
    • Prints text or JSON, and --fail-on-honor-only turns it into a CI gate (exit 1)
    • Never edits your project, never runs your hooks, makes no network calls

    Honest limits, stated in the README: a match means something mentions the rule by name, not that the hook works. Read the hooks you rely on.

    What is inside

    • SKILL.md, README.md, LICENSE
    • scripts/rule_enforcement_map.py
    • tests/test_package.py (17 tests)
    • examples/sample_project/ (the project used in the example above)

    Requirements and safety

    Python 3.8+, macOS or Linux, standard library only. Read-only: no installers, no background services, no credentials, no network.

    License

    Personal, non-transferable license for your own projects, including commercial work. No resale or redistribution.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Code Quality & Review.

    Matches filenames only, does not verify hook logic. Ignores filenames < 4 chars. Skips files > 512 KB. Requires Python 3.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 5 days ago

    Software
    Python 3.8+Git

    Permissions

    Terminal / Shell
    Read Files
    Listed5 days ago

    What's inside

    Frequently Asked Questions

    It shows which rule files in .claude/rules actually have something behind them (a hook, a script a hook runs, a git hook, a CI job or a test that mentions the rule by name) and which depend on the agent remembering them. It also flags rules that declare an enforcer path that does not exist.

    The scanner is a plain Python script that reads .claude/rules/*.md and looks in Claude Code's usual places (settings.json hooks, .claude/hooks, git hooks, CI workflows, folders you name), so it is built for Claude Code projects. The package follows the SKILL.md format.

    Yes. Run python3 scripts/rule_enforcement_map.py --fail-on-honor-only as a step in any CI job. It exits 1 when a rule is honor-system or has a broken claim. It needs Python 3.8+ and makes no network calls.

    It does not judge whether a rule is stale. A rule counts as enforced when something mentions it by name, or by keywords you declare with an enforced-by comment or enforce_keywords in the front matter. Everything else is listed as honor-system, and you decide whether that is acceptable.

    SKILL.md, README.md, LICENSE, scripts/rule_enforcement_map.py, 17 tests, and the examples/sample_project used in the example. Personal, non-transferable license for your own projects, including commercial work; no resale or redistribution.

    Yes. A match means something mentions the rule by name, not that the hook works. The tool never runs your hooks and does not check them for correctness, so read the hooks you rely on. It needs Python 3.8+ on macOS or Linux and uses only the standard library.

    New in Code Quality & Review