Content
78%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.
A well-structured, actionable body for a simple generative skill: concrete CLI commands, a curl fallback, and clean sectioning. The main gaps are the absence of a full example invocation for the primary CLI and no output-verification checkpoint.
Suggestions
Show one copy-paste-ready `muapi image generate` invocation with actual argument syntax instead of only naming the command and parameters.
Clarify the aside 'Use the plan if generating more than one variant' — it references a plan that is not defined anywhere in the skill.
Add a brief checkpoint after Phase A (e.g. confirm the image request succeeded via `muapi predict wait` before composing the caption).
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is efficient — an inputs table, two tight phases, and brief operational notes with no explanations of concepts Claude already knows. It is not a 5 due to minor trimmable content: the marketing-tone Note 'Prioritize scroll-stopping first impressions — the image must visually communicate the brief within 2 seconds' and the confusing aside 'Use the plan if generating more than one variant'. | 4 / 5 |
Actionability | Mostly executable guidance: a concrete command ('Call `muapi image generate` (model=nano-banana-2, aspect_ratio=`{{format}}`)') plus a complete curl fallback with headers and a polling command. It is not a 5 because the primary CLI invocation is named but not shown as a copy-paste-ready example with full argument syntax, and the image-prompt step is guidance rather than a concrete worked example. | 4 / 5 |
Workflow Clarity | Clear two-phase sequence with explicit ordering ('After the image is generated'), conditional branching (edit vs. generate), and failure-handling guidance (auth setup, polling). It is not a 5 because there are no explicit validation checkpoints (e.g. verifying the image generated successfully before composing the caption); the destructive/batch cap does not apply since this is a generative, single-output skill. | 4 / 5 |
Progressive Disclosure | There are no bundle files and the body is ~50 lines with well-organized sections (Inputs, Steps, Notes, Trigger Keywords, agent notes), so the simple-skill exception applies and progressive disclosure scores 5 on well-organized sections alone. Nothing that belongs in a separate file is inlined. | 5 / 5 |
Total | 17 / 20 Passed |