Content
93%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 single-file skill body: terse, concrete, and immediately executable, with a well-sequenced workflow and a built-in self-check. The only weakness is that the self-check checklist has no explicit error-recovery loop, which caps workflow clarity at 4.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Lean and efficient throughout — e.g. 'Charts: inline SVG only, no JS libraries. A line chart is ~10 lines of <polyline> with a subtle area fill.' It assumes Claude's competence (no explanation of what CSS Grid or semantic HTML are) and every instruction earns its place, matching the top anchor. | 5 / 5 |
Actionability | Fully concrete, executable guidance: pixel ranges ('Left sidebar (220–260px)'), counts ('6–8 nav links', '3–4 KPI cards'), a copy-paste-ready <artifact> output block, and a specific self-check checklist. For an instruction-only skill this matches the top anchor. | 5 / 5 |
Workflow Clarity | A clear 6-step sequence (Read DESIGN.md → Classify → Lay out → Write → Charts → Self-check) with a terminal validation checklist ('Every color comes from DESIGN.md tokens', 'Accent used at most twice'). It lacks an explicit fix-and-retry feedback loop around the self-check, so it fits the 'most checkpoints, minor validation gaps' anchor rather than 5. | 4 / 5 |
Progressive Disclosure | The body is under 50 lines, organized into clear sections (## Workflow, ## Output contract), and appropriately references injected context (the active DESIGN.md) rather than inlining it. No bundle files exist or are needed, so per the simple-skill guideline the well-organized single file matches the top anchor. | 5 / 5 |
Total | 19 / 20 Passed |