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 instruction-only skill body: maximal information density, concrete before/after examples in the target register, explicit exception handling, and a pre-send checklist that serves as a validation step. There is no wasted content and no missing structure for a skill of this scope.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~54-line body is pure rules and examples with zero padding — no concept explanations Claude already knows, and even meta-guidance ("Token savings are small and noisy; never claim them") earns its place. Every token earns its place. | 5 / 5 |
Actionability | Concrete Bad/Good example pairs in real 文言 (e.g. "每繪新生對象參照,故重繪;以 `useMemo` 包之則免"), explicit floors (negation characters, Arabic numerals, verbatim identifiers), and a literal pre-send checklist give fully actionable guidance; per the rubric's code-vs-instruction note, an instruction-only skill needs no code when guidance is this concrete. | 5 / 5 |
Workflow Clarity | A single-action skill whose action is unambiguous: persistence semantics ("Unsure? Still on."), status-command behavior, escape hatches ("When to break the rules"), and a pre-send validation check. It is not a destructive or batch skill, so no validation cap applies and the simple-skill exception holds. | 5 / 5 |
Progressive Disclosure | Well-organized single-file skill with clear sections (Persistence, Floor, Rules, When to break the rules, Pre-send check); no bundle files exist and nothing is inlined that belongs in a separate file, matching the rubric's simple-skill guidance for short skills needing no external references. | 5 / 5 |
Total | 20 / 20 Passed |