Content
96%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The content is an exemplar of lean, actionable, well-structured operational guidance: executable commands, a precise output contract, and clean sectioning with implementation pushed into a bundled script. The only notable gap is the absolute, machine-specific path used to invoke the bundled script.
Suggestions
Reference the bundled script via a relative path (e.g. scripts/clawsweeper-status.sh) instead of the hardcoded /Users/steipete/... absolute path so the skill is portable.
Briefly note the presence of the test script and fixture as supporting bundle files so reviewers/users can discover them.
Consider adding a one-line note on what to do if the script exits non-zero (e.g. missing gh auth), since it is the single entry point.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and operational with no concept-explaining padding, assumes Claude's competence, and nearly every token (script flags, section contract, interpretation rules) earns its place. | 5 / 5 |
Actionability | It provides copy-paste-ready executable invocations with real flags (--hours, --repo, --limit) plus a concrete output contract with exact section names and prescribed phrasing, covering the common cases. | 5 / 5 |
Workflow Clarity | This is a single-action read-only skill (run the script, then report sections) with an unambiguous sequence and explicit reporting rules; no destructive/batch operation is involved, so the simple-skill exception applies and no validation loop is required. | 5 / 5 |
Progressive Disclosure | The SKILL.md is a concise overview that delegates implementation to the bundled script (one level deep) with well-organized sections, but it references the script via a machine-specific absolute path rather than a relative bundle path, a minor navigation/portability gap. | 4 / 5 |
Total | 19 / 20 Passed |