Content
93%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.
Lean, fully executable single-purpose skill with a clear branching workflow and well-organized sections. The only gap is the lack of an explicit verification step after the destructive removal, though the existence pre-check mitigates the risk.
Suggestions
Add a post-deletion verification step (e.g., confirm `.cursor/ralph` no longer exists) before reporting cancellation, since `rm -rf` is destructive.
Consider capturing and reporting the iteration count read in step 3 inside the final report message to make the output deterministic rather than templated as "iteration N".
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is ~25 lines with no padding or explanation of concepts Claude already knows; every section (Trigger, Workflow, Output) earns its place. | 5 / 5 |
Actionability | It gives a copy-paste-ready command (`rm -rf .cursor/ralph`), a concrete file path, a specific field to read (`iteration:`), and exact report messages for both branches. | 5 / 5 |
Workflow Clarity | A clear sequenced workflow with an explicit existence-check branching checkpoint; the destructive `rm -rf` is guarded by the pre-check, but there is no explicit post-deletion verification step, leaving a minor validation gap. | 4 / 5 |
Progressive Disclosure | A simple skill under 50 lines with no need for external references, organized into clear Trigger/Workflow/Output sections, which satisfies the simple-skill exception. | 5 / 5 |
Total | 19 / 20 Passed |