Content
75%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, mostly executable workflow with a clear 5-step sequence, a decision table for actor selection, and concrete error recovery guidance. Remaining gaps are the placeholder input JSON with no example, generic boilerplate sections, and the absence of explicit validation feedback loops around batch scraping runs.
Suggestions
Add a concrete example JSON input for at least one actor so the `--input 'JSON_INPUT'` placeholder is grounded.
Delete or rewrite the generic "When to Use" and "Limitations" sections, which restate template text and add no skill-specific information.
Add a verification step after Step 4 (e.g., check the output file is non-empty and spot-check a few rows) to close the validation gap for batch runs.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is efficient — a compact actor decision table, copy-paste commands, and an error table with no concept explanations. Not a 5 because the boilerplate "When to Use" section ("Use this skill when tackling tasks related to its primary domain or functionality as described above") and generic Limitations section are template filler that could be trimmed. | 4 / 5 |
Actionability | Provides executable commands (the mcpc fetch call with a concrete example actor, and three complete `node --env-file=.env ... run_actor.js` invocations) plus concrete error fixes. Not a 5 because `JSON_INPUT` is a placeholder with no example input object, and the referenced script (`reference/scripts/run_actor.js`) is not in the bundle for verification. | 4 / 5 |
Workflow Clarity | A copyable progress checklist and five clearly sequenced steps, with a schema-fetch checkpoint (Step 2) and a user-preference gate (Step 3) before any run, plus an error-handling table with recovery actions. Not a 5 because error handling is a lookup table rather than embedded validate→fix→retry loops, and output verification beyond "report number of results" is thin for batch scraping runs. | 4 / 5 |
Progressive Disclosure | Well-organized sections with the only external reference being a single script path (`${CLAUDE_PLUGIN_ROOT}/reference/scripts/run_actor.js`), one level deep; no bundle files exist to score against. Not a 5 because the skill exceeds 50 lines and the 19-row actor catalog is inlined content that could live in a reference file. | 4 / 5 |
Total | 16 / 20 Passed |