Content
77%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is highly actionable with executable examples and a clear, validated generate-verify-retry workflow. It loses points for verbosity from redundant verification sections and a monolithic structure that could benefit from splitting peripheral guidance into reference files.
Suggestions
De-duplicate the post-generation verification guidance: keep one authoritative '生成后必须验证' section and cross-reference it instead of restating it.
Move the long video-cover/poster best-practices and model-开通 onboarding sections into a separate reference file (e.g. references/best-practices.md) and link to it from SKILL.md to reduce the monolithic wall of text.
Trim the Pitfalls section of details already covered by the Parameters table to avoid restating the same constraints twice.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Mostly efficient and assumes Claude's competence on API-specific quirks (size limits, fallback, watermark), but repeats the post-generation verification guidance in two overlapping sections and re-states parameter info in Pitfalls, so it could be tightened. | 2 / 3 |
Actionability | Provides concrete, copy-paste-ready bash commands via the wrapper, a complete parameter table, multiple worked examples, and a full end-to-end invocation — fully executable guidance. | 3 / 3 |
Workflow Clarity | Sequences the generate→verify→retry flow with an explicit mandatory validation checkpoint ("每张图生成后必须用 image 工具检查"), a bounded retry loop (max 3), and clear failure/escalation handling. | 3 / 3 |
Progressive Disclosure | No nested/deep references (good — single level, inline), but the ~250-line file is monolithic; the video-cover best-practices and model-开通 guidance sections could be split into separate reference files for better navigation. | 2 / 3 |
Total | 10 / 12 Passed |