Content
100%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.
An exemplary short workflow skill: concrete commands and label names, an explicit repeat-until-green validation loop, a human-in-the-loop checkpoint, and tight authorization scoping, with zero padding. Structure and length are exactly right for a single-purpose slash-command skill.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The 28-line body is lean and assumes Claude's competence: every sentence adds non-obvious information such as the authorization scope ('It does not authorize a merge'), Greptile's automatic review, and why post-approval changes must be reported ('the user has not seen those'). Nothing explains concepts Claude already knows. | 5 / 5 |
Actionability | Guidance is fully executable for an instruction-only skill: exact commands ('gh pr ready <number>', 'gh pr edit <number> --add-reviewer <login>'), concrete label names (ci:docs, ci:focus, ci:normal, ci:daily), and explicit delegation to named sibling skills for the fix/wait loops. The only open-ended directive ('find the GitHub login') is a reasonable assumption of competence. | 5 / 5 |
Workflow Clarity | The five steps are clearly sequenced with an explicit validation feedback loop ('Repeat until each check on the head commit passed, Greptile reviewed the head commit, and no bot thread is unresolved') plus a human-decision checkpoint ('present it and wait: request no review before the answer'), and the 'After the request' section extends the loop to reviewer comments. | 5 / 5 |
Progressive Disclosure | This is a simple sub-50-line skill with no bundle files and no need for external references; sections are well organized and the sibling-skill links (../build-it, ../check-pr, ../handle-pr-comments) are one level deep and clearly signaled, so the simple-skill exception applies. | 5 / 5 |
Total | 20 / 20 Passed |