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, largely executable skill with a clear five-step workflow and useful error-recovery guidance. The main gaps are the missing output-validation checkpoint for batch scraping runs and the undefined JSON_INPUT placeholder, plus an oversized inline Actor catalog that belongs in a separate reference file.
Suggestions
Add an explicit verification step after running the script (e.g. check the output file exists and is non-empty, and confirm the row count matches the requested number of results) before summarizing findings.
Provide one concrete example of a JSON_INPUT value (e.g. '{"searchStrings": ["Competitor Name"], "location": "New York", "maxCrawledPlaces": 10}') so the run commands are copy-paste ready.
Move the 27-row Actor selection table into a references file (e.g. references/actors.md) and keep a short decision guide inline, trimming filler like '(No need to check it upfront)'.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly lean: a dense actor-selection table, copy-shaped bash commands, and a compact error-handling list with no explanations of concepts Claude already knows. Minor padding keeps it from a 5 — the odd parenthetical "(No need to check it upfront)" and the vague "Number of results: Based on character of use case" could be trimmed or made concrete. | 4 / 5 |
Actionability | Guidance is mostly executable: exact bash for fetching the Actor schema via mcpc, and per-format node invocations with real flags (--actor, --input, --output, --format). It falls short of anchor 5 because JSON_INPUT is never exemplified — no sample input object is shown — and the referenced run_actor.js arguments are undocumented in the bundle; it stays above anchor 3 because the commands are concrete, not pseudocode. | 4 / 5 |
Workflow Clarity | The five-step sequence is clearly ordered with a copyable progress checklist and an error-handling section mapping failures to fixes. However, the workflow drives batch scraping operations that consume paid API runs and produce data files, yet includes no validation/verification checkpoint on outputs (e.g. confirm the CSV is non-empty or row counts match expectations) — per the rubric's batch-operation rule this caps workflow clarity at 3 even though the error-recovery guidance is otherwise good. | 3 / 5 |
Progressive Disclosure | Section structure is good: prerequisites, a checklist-driven workflow, and error handling, with a single clearly-signaled external script reference (${CLAUDE_PLUGIN_ROOT}/reference/scripts/run_actor.js) that is one level deep. It is not a 5 because the 27-row Actor catalog is reference-like data inlined in SKILL.md that could be split into a references file, and no bundle reference/script files exist to organize that material. | 4 / 5 |
Total | 15 / 20 Passed |