Content
36%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 functions as a thin catalogue pointer: it is reasonably concise and sectioned, but provides almost no executable guidance or concrete workflow, redirecting the real instructions entirely to an external URL.
Suggestions
Inline at least one concrete, runnable example (e.g., a minimal Python/SVG snippet that renders a poster to PNG) so the skill is actionable without an internet hop.
Replace the external-redirect 'How to use' steps with an explicit local install-and-invoke sequence with a verification step (e.g., confirm the skill appears in the agent's skill list before invoking).
Bundle key references locally (e.g., a REFERENCES.md with design principles) and link them one level deep instead of pointing only at the upstream GitHub URL.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is short and avoids explaining concepts Claude already knows, but it restates the description verbatim under 'What it does' and includes mild filler ('Curated from Anthropic's official skills repository') that could be trimmed. | 3 / 5 |
Actionability | The only concrete command is `open <url>`, which launches a browser rather than executing the skill's task; there is no executable code or specific steps for actually producing art. | 2 / 5 |
Workflow Clarity | A rough inspect→install→invoke sequence is present but each step defers to an external README for specifics and there is no validation, leaving the steps poorly defined with major gaps. | 2 / 5 |
Progressive Disclosure | The single file has clear section headers, but it offloads all real content to an external GitHub URL rather than curating local one-level-deep references, so structure is only partially effective. | 3 / 5 |
Total | 10 / 20 Passed |