assaylisted

Rules quality auditor for existing Claude Code instruction files (CLAUDE.md, .claude/rules/). Grades, scores, audits, reviews, and critiques the structural clarity of existing rules. Reads and evaluates only — never creates new rules (that's forge), never reviews code or PRs, and ignores non-Claude-Code config like eslint or prettier.
V-Songbird/claude-plugins · ★ 0 · Code & Development · score 78

Install: claude install-skill V-Songbird/claude-plugins

# Rules Assay — Quality Scoring Score the project's Claude Code instruction files for structural clarity — how clearly Claude can parse and apply each rule. **Language:** Scoring is English-only. Non-English rules will receive inaccurate scores. If arguments are provided, treat them as flags: $ARGUMENTS Supported flags: `--fix` (suggest + apply rewrites), `--verbose` (detailed factor breakdown), `--json` (machine-readable) **Architecture**: Four factors (F1, F2, F4, F7) are scored by deterministic Python scripts. Score two judgment factors (F3, F8) plus edge-case patches. The `run_audit.py` orchestrator handles all pipeline mechanics — the only bash interactions are calling its modes and writing judgment data files. ## User interaction discipline — read before any phase Every point in this skill where the user is asked to pick between options, approve an action, or confirm next steps **must** use the `AskUserQuestion` tool. Never ask a decision question as free-form text in chat — users may not notice the question, and there is no way for the harness to chain into the next step without the structured selection. Rules: 1. **If you are asking the user to choose between options, use `AskUserQuestion`.** Plain-text questions like "Want me to do X, or Y?" break the UX. 2. **If the choices are not mutually exclusive** (e.g. multiple coverage gaps, multiple files to split), set `multiSelect: true` on that question. 3. **If you finish a step and have a new finding to offer**, tr