Content
85%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 tight, executable recipe: exact prompt templates per shot type, explicit model selection logic, and a raw-endpoint fallback make it directly actionable. The only real flaws are minor duplication and unspecified CLI argument syntax.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean — a compact inputs table, one generation step, and brief notes — with no explanations of concepts Claude already knows. Not 5 because the description is restated verbatim as a bold line under the H1 and the trigger-keywords section largely duplicates routing information. | 4 / 5 |
Actionability | Concrete commands throughout: named CLI models ('veo3.1-text-to-video', 'veo3.1-image-to-video'), literal prompt templates for all four shot types, and a copy-paste curl fallback with polling. Not 5 because the exact argument syntax for 'muapi video generate' (how prompt and aspect_ratio are passed) is never shown, leaving a minor gap. | 4 / 5 |
Workflow Clarity | This is a simple single-purpose skill and the flow is unambiguous: gather inputs, branch on the presence of 'reference_image', build the prompt from the shot_type table, submit, then offer follow-ups — with auth setup and a polling fallback explicitly handled in the agent notes. No destructive or batch operations exist to require validation checkpoints, so the simple-skill exception applies. | 5 / 5 |
Progressive Disclosure | The body is well under 50 lines, has no external reference files, and no content that needs splitting out — inputs, workflow, notes, and agent instructions are cleanly sectioned. Per the guideline, a short skill with no need for external references and well-organized sections scores 5. | 5 / 5 |
Total | 18 / 20 Passed |