Content
61%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 skill body is well-organized and gives mostly concrete, executable guidance centered on a clear two-step workflow, but it spends tokens re-explaining design concepts Claude already knows and lacks any validation or error-recovery checkpoint around image generation. The dangling script references also slightly undermine its structure.
Suggestions
Add a validation step after script invocation, e.g., check the generated image against the Constraints section (no hands/devices, contrast, consistent atoms) and regenerate with adjusted flags if it fails — this addresses the weakest dimension, workflow clarity.
Trim the "Core Competencies" section and generic design-token explanations (Atomic Design, Nielsen's heuristics, 8pt grid, WCAG basics) down to one-line pointers; keep only muapi-specific and non-obvious guidance to improve conciseness.
Either include the referenced "scripts/generate-mockup.sh" (and document its full flag set and outputs) or remove/fix the dangling references so the documented invocation actually resolves in the bundle.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly task-relevant, but the "Core Competencies" section, the Nielsen's Heuristics mention, and the generic design-token explanations (Inter/Roboto, 8pt grid, WCAG) re-teach design concepts Claude already knows, and the intro sentence duplicates the header line — more than minor trimmable padding, so anchor 3 rather than 4. | 3 / 5 |
Actionability | The bash invocation is concrete and executable-shaped ("bash scripts/generate-mockup.sh --desc ... --platform mobile --theme dark --style glassmorphism") and the numbered prompt-expansion steps give specific directives ("Input fields with 4px radius", "Inter Sans-serif"). Minor gaps: the script's outputs and full flag set are undocumented, and both referenced scripts are missing from the bundle — mostly executable, not fully copy-paste ready, so anchor 4. | 4 / 5 |
Workflow Clarity | A real sequence exists ("Step 1: Define the Product Brief" → "Step 2: Invoke the Script", preceded by the Prompt Optimization Protocol), but there are no validation checkpoints or error-recovery loops — e.g., no step to verify the generated image meets the "no hands/phones" and contrast constraints before delivering. Anchor 3: steps present, checkpoints missing. | 3 / 5 |
Progressive Disclosure | The body is well-sectioned and self-contained (Intent Mapping, Tokens, Protocol, Constraints), but no bundle files exist while the body references "scripts/generate-mockup.sh" and "core/media/generate-image.sh" that aren't present, and the ~75-line body exceeds the 50-line simple-skill threshold while inlining reference-style tables. Good structure with minor organization gaps — anchor 4. | 4 / 5 |
Total | 14 / 20 Passed |