Content
72%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a well-structured, concise skill that clearly communicates what to validate and how to report results. Its main weakness is that the validation process relies on Claude inferring implementation details (regex patterns, YAML parsing, line counting) rather than providing executable validation code or a script. The workflow could also benefit from explicit error-handling guidance for batch validation.
Suggestions
Add executable code snippets for key validation checks (e.g., a regex for name format validation, a command to count body lines, YAML frontmatter parsing)
Add a feedback loop step: after reporting failures, instruct to fix issues and re-run validation until all checks pass
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The content is lean and efficient. It uses tables for compact rule presentation, avoids explaining what skills are or how Claude Code works, and every section serves a clear purpose. No unnecessary padding or concept explanations. | 3 / 3 |
Actionability | Provides a concrete checklist and output format, plus a bash command for discovery. However, the actual validation logic is left implicit — there's no executable script or code to parse YAML frontmatter, check name formats with regex, or count lines. Claude must infer the implementation from the checklist. | 2 / 3 |
Workflow Clarity | The three-step workflow (find directories, check rules, report) is clear in sequence, but there are no validation checkpoints or feedback loops. For a batch operation across multiple skills, there's no guidance on what to do when failures are found, no prioritization of errors, and no explicit 'fix and re-validate' loop. | 2 / 3 |
Progressive Disclosure | For a skill under 50 lines with no need for external bundle files, the content is well-organized with clear sections (checklist, how to run, references). External references to the spec and best practices are one-level deep and clearly signaled with markdown links. | 3 / 3 |
Total | 10 / 12 Passed |