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.
A highly actionable, well-structured skill: concrete commands, complete JSON examples, and genuinely useful operational caveats (charges, pagination limits, run reconciliation) with almost no filler. Its main weaknesses are the missing 'Use when' guidance in the description and monolithic inline reference data (marketplace lists, per-workflow schemas) that belongs in a separate reference file.
Suggestions
Move the exhaustive marketplace domain lists and the per-workflow input schema tables into a one-level-deep reference file (e.g. references/marketplaces.md), keeping only a few representative examples inline in SKILL.md.
Add an explicit validation checkpoint between running the script and summarizing results — e.g. verify the item count and spot-check exported records — to complete the feedback loop for this batch, billable operation.
Clarify the script location convention (reference/scripts/ vs references/scripts/) and confirm the referenced run_actor.js ships with the skill bundle.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is efficient — tight tables, JSON snippets, and an error/solution table with no explanations of concepts Claude already knows — with only minor trimmable padding such as the "Best for" audience lines and the ~60 inline marketplace domain names. | 4 / 5 |
Actionability | Bash commands are copy-paste ready with exact flags (--actor, --input, --output, --format), each workflow has complete executable JSON input examples, output fields and sort options are enumerated, and the error table maps concrete failures to solutions, covering the common cases. | 5 / 5 |
Workflow Clarity | A progress-tracking checklist, a workflow-selection table, and Steps 1–3 give a clear sequence, and the error-handling table plus "keep its run ID, inspect/reconcile status" guidance provide most checkpoints; the gap is no explicit verify-output checkpoint before summarizing, which keeps it below anchor 5. | 4 / 5 |
Progressive Disclosure | Headers organize the content well and the script path is clearly signaled, but no bundle files exist, and bulk reference material (60+ marketplace domains, full per-workflow input schemas) is inlined in SKILL.md instead of being split into a one-level-deep reference file. | 3 / 5 |
Total | 16 / 20 Passed |