Content
86%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 lean, instruction-only skill: every bullet is a specific, testable constraint, mode semantics (canon/AU/OOC/CP) are defined once, and the scale needs no external files. The only soft spot is a few abstract directives without illustrative examples of desired voice or output.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Every bullet earns its place ('Preserve recognizable voice through motive, knowledge, rhythm, and behavioral choices, not copied phrases'); no padding and no explanation of concepts Claude already knows — the AU/OOC/CP mode definitions are domain knowledge worth stating. | 5 / 5 |
Actionability | Concrete, testable decision rules ('keep the gap open or ask; do not turn a guess into canon'; new characters 'must serve the new line and remain distinguishable from source canon'), but some lines are abstract ('Build a new dramatic line rather than replaying source scenes') and no illustrative examples are given — minor gaps, not missing key details. | 4 / 5 |
Workflow Clarity | A flat, single-purpose constraint list with no destructive or batch operations needing validation; the single action is largely unambiguous, but minor process gaps remain (e.g., what to do when the user does not name a mode, how to verify voice fidelity). | 4 / 5 |
Progressive Disclosure | Six lines total with no references/, scripts/, or assets/ bundle; nothing here warrants splitting into separate files, and the inline bullet list is appropriately organized at this scale, so the simple-skill exception applies. | 5 / 5 |
Total | 18 / 20 Passed |