Content
75%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.
The body is an efficient, well-structured, and highly actionable strategy reference for an instruction-only skill — no code absence penalty applies since the guidance is concrete and specific. Its main gaps are minor: a couple of filler sections, no worked example, and no progressive split of detailed material into reference files.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is a dense, checklist-style reference with concrete heuristics and no explanations of concepts Claude already knows. Minor trimmable filler exists (the vacuous "When to Use" section saying only "This skill is applicable to execute the workflow or actions described in the overview" and the "You are an expert" roleplay preamble), keeping it just below the lean-efficiency anchor of a 5. | 4 / 5 |
Actionability | Concrete, specific guidance throughout: numeric heuristics (30-60 second triggers, 25-50% scroll depth, 2-5% benchmarks, 7-30 day cooldowns), templated copy patterns ("Get [result] in [timeframe]"), a mandatory close-behavior spec, and a required output format. It is not a 5 because no fully worked end-to-end example of a complete popup recommendation is provided — guidance is directional rather than copy-paste complete. | 4 / 5 |
Workflow Clarity | A clear required sequence is present: Initial Assessment (Required) establishes context first, then trigger/type/copy/design/frequency decisions, compliance checks, and a required Output Format acting as a completion checklist, supplemented by Questions to Ask. This is advisory (non-destructive, non-batch) work, so the missing-validation cap does not apply; it stays at 4 rather than 5 because no verification or feedback step (e.g., how to sanity-check a recommendation against benchmarks) exists. | 4 / 5 |
Progressive Disclosure | Scored against the actual bundle structure: no reference files exist, and the single ~345-line body is well organized into twelve numbered sections with clear headers and cross-references to sibling skills (form-cro, page-cro) for boundary navigation. It is not a 5 because at this length some self-contained sections (copy framework examples, compliance details, popup type requirements) could be split into one-level-deep reference files to keep SKILL.md a leaner overview. | 4 / 5 |
Total | 16 / 20 Passed |