Content
71%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.
A lean, appropriately scoped skill for a single review task with a clear workflow and an effective external-reference design. Its weaknesses are a fully duplicated step list and the absence of any error handling around the network fetch.
Suggestions
Merge the "Usage" section into "How It Works" (or replace it with only the delta — the no-argument behavior) to remove the duplicated four-step list.
Add a one-line fallback for when the guidelines URL is unreachable (e.g., inform the user or use a cached copy if available).
Specify how findings should be grouped or prioritized when reviewing multiple files, since the body only defines the per-finding `file:line` format.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is short and free of concept explanations, but the "Usage" section repeats the four steps of "How It Works" almost verbatim, so the content could clearly be tightened by removing the duplicate. | 3 / 5 |
Actionability | Concrete guidance is present — an exact source URL, a named tool ("Use WebFetch"), and a specific output format ("terse `file:line` format") — with only minor gaps such as no fallback when the fetch fails. | 4 / 5 |
Workflow Clarity | A clear four-step sequence (fetch, read, apply rules, output findings) with a concrete output-format checkpoint; no error-recovery loop for a failed guidelines fetch keeps it below a 5, and the review is read-only so the destructive-operation cap does not apply. | 4 / 5 |
Progressive Disclosure | The skill is under 50 lines, single-purpose, and has no local bundle files; sections are well organized and the only reference (the guidelines URL) is clearly signaled and one level deep, so the simple-skill exception applies. | 5 / 5 |
Total | 16 / 20 Passed |