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, well-structured wrapper skill that appropriately delegates the rule catalog to a fetched external source and keeps the SKILL.md body minimal. Its weaknesses are a fully duplicated step list across two sections and the absence of any handling for fetch failure or an output-format example.
Suggestions
Merge the duplicated four-step lists in "How It Works" and "Usage" into a single workflow section to remove the redundancy penalized in conciseness.
Add a fallback for fetch failure (e.g., retry once, then inform the user the guidelines are unavailable rather than proceeding without rules), which would also strengthen workflow checkpoints.
Include a one-line example of the expected `file:line` finding format so the output contract is concrete without fetching the guidelines first.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body has no concept over-explanation, but the "How It Works" steps ("Fetch the latest guidelines... Read the specified files... Check against all rules... Output findings") are repeated almost verbatim in the "Usage" section ("Fetch guidelines... Read the specified files... Apply all rules... Output findings"), a full duplicated section that should be collapsed into one. This matches the "mostly efficient but could be tightened" anchor rather than the minor-trim anchor 4. | 3 / 5 |
Actionability | The skill supplies a copy-paste-ready URL, names the exact tool ("Use WebFetch to retrieve the latest rules"), specifies a concrete output contract ("the terse `file:line` format"), and defines a fallback ("If no files specified, ask the user"), which is mostly executable guidance. It does not reach anchor 5 because no example of the expected output format is shown and per-rule application is left abstract to the fetched content. | 4 / 5 |
Workflow Clarity | The steps are numbered, clearly sequenced (fetch → read → check → output), and include a fallback for missing arguments, matching anchor 4's "clear sequence with most checkpoints present, minor validation gaps". There is no checkpoint for fetch failure (offline or moved URL) and no error-recovery loop, so it does not reach anchor 5; the operation is read-only, so the destructive/batch cap to 3 does not apply. | 4 / 5 |
Progressive Disclosure | The skill is under 50 lines with no bundle files needed and well-organized sections (How It Works, Guidelines Source, Usage), while the bulk content (the rules themselves) correctly lives one level deep at the fetched external URL for freshness. Per the rubric's simple-skill exception this warrants anchor 5. | 5 / 5 |
Total | 16 / 20 Passed |