Content
68%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 that routes cleanly via a decision tree and gives concrete parameter tables and worked workflows. The main weakness is missing validation/verification checkpoints in the batch and conversion workflows.
Suggestions
Add an explicit verify step to the batch conversion workflow, e.g. after conversion call ImageProcess(action: "get-info") on one output to confirm the new format before reporting success.
Replace "…" path placeholders in the worked workflow examples with representative concrete file paths so the snippets are fully copy-paste ready.
Add a brief validation checkpoint to the download→extract-audio workflow (confirm the downloaded filePath exists before invoking VideoConvert).
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and assumes Claude's competence — it does not explain what sharp/FFmpeg are — with only minor spots (e.g. the WebP vs JPEG size note) that could be trimmed. | 4 / 5 |
Actionability | Concrete action tables with key parameters and executable workflow snippets (e.g. ImageProcess(action: "convert", filePath: "…", format: "webp", outputPath: "…")) cover common cases, but "…" placeholders keep examples just short of copy-paste ready. | 4 / 5 |
Workflow Clarity | A decision tree and sequenced workflows are present, but the batch conversion workflow ("对每张图片调用 ImageProcess(action: "convert", format: "webp")") has no output verification step, so workflow clarity is capped at 3 per the batch/destructive-operations guideline. | 3 / 5 |
Progressive Disclosure | A single self-contained SKILL.md with well-organized sections (工具清单, 决策树, per-tool sections, 常见工作流, 常见错误) and no nested references; no bundle files exist, so organization is appropriate with only minor structure gaps. | 4 / 5 |
Total | 15 / 20 Passed |