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.
The body is a well-structured, actionable recipe: a clear inputs table, an explicit generation step with model IDs, prompt template, conditional logic, and practical platform notes. Its few gaps are the missing literal CLI command syntax and the absence of a validation checkpoint for the generated cover.
Suggestions
Show one complete example invocation of `muapi image generate` with actual flags (prompt, model, aspect ratio) so the command is copy-paste ready.
Add a post-generation validation step, e.g. check the image dimensions/aspect ratio and confirm the text overlay rendered correctly before presenting it, with a retry on failure.
Remove the verbatim repetition of the frontmatter description as the bold line under the title to save tokens.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean with no concept explanations Claude already knows (no 'what is Xiaohongshu' padding), but the frontmatter description is repeated verbatim as a bold line under the title and trigger keywords duplicate description terms — minor instances that could be trimmed, matching the 'efficient; minor over-explanation' anchor. | 4 / 5 |
Actionability | Concrete guidance throughout: named CLI commands ('muapi image edit', model='bytedance-seedream-v4.5-edit'), a full prompt template with {{placeholder}} substitution rules, aspect-ratio mapping, and a curl fallback with polling. It falls short of anchor 5 only because the exact CLI invocation syntax (flags) is never shown, so commands are not fully copy-paste ready. | 4 / 5 |
Workflow Clarity | A clear single-phase sequence with explicit conditional branching (reference_image → edit vs. generate), an auth prerequisite, aspect-ratio logic, and follow-on offers. Not anchor 5 because there is no checkpoint verifying the generated image or that the text overlay rendered correctly — a real failure mode for image text rendering. | 4 / 5 |
Progressive Disclosure | This is a simple ~50-line, single-purpose skill with no need for external references (no references/, scripts/, or assets/ bundle exists, and none is required), and the body is cleanly organized into Inputs, Steps, Notes, Trigger Keywords, and Executing Agent sections — per the simple-skill exception this earns the top score. | 5 / 5 |
Total | 17 / 20 Passed |