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 minimal skill body: executable commands for all common invocation modes, precise exit-code documentation with a remediation loop for the OCR failure case, and guidance on when to prefer libraries over the CLI — all in a lean, well-organized form with no filler.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and efficient with zero padding — every line teaches CLI-specific knowledge Claude cannot know (exit codes, 'The CLI never prompts', 'Rerun with --ocr hosted', stdin mode), and no concepts Claude already knows are explained, matching the score-5 anchor. | 5 / 5 |
Actionability | Three copy-paste-ready commands cover the common cases (stdout, file output via -o, stdin with --format csv), plus concrete exit-code semantics (0/1/2/3), an exact remediation command ('--ocr hosted'), and named library packages per ecosystem — fully executable guidance matching the score-5 anchor. | 5 / 5 |
Workflow Clarity | This is a simple single-purpose skill where the single action (run the anydoc CLI) is unambiguous, satisfying the simple-skill exception; it additionally provides explicit validation via exit codes and an error-recovery path (exit 3 → rerun with --ocr hosted), exceeding the unambiguous-single-action bar. | 5 / 5 |
Progressive Disclosure | The skill is under 50 lines with no need for external references, and the content is well organized (usage commands block followed by a numbered rules list), which the rubric's guidance explicitly scores as 5 for simple skills with well-organized sections. | 5 / 5 |
Total | 20 / 20 Passed |