Content
96%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.
An exceptionally dense, operational skill body: exact tool usage, concrete quantities, precise merge/counting rules, cost and failure handling at every paid step, and a fully specified deliverable format. The only structural note is that everything lives inline in a ~95-line file with no reference split.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | There is no concept explanation or padding; every section carries operational specifics, e.g. "merge the prompts from all calls by their normalized text (lowercase, trimmed, collapsed whitespace)" and "`www.semrush.com`, `semrush.com` and `sv.semrush.com` are one domain". The few rationale lines ("That source mix is the useful finding") directly guide reporting behavior, so every token earns its place. | 5 / 5 |
Actionability | Fully executable for an instruction-only skill: exact tool names and parameters (`research_ai_visibility_prompts`, `explore_prompt` "with its most representative prompt, the default ChatGPT model and `highlightBrand`"), a copy-paste `update_project_context` payload with `{ updates: [{ appendResearchLog: ... }] }`, concrete counts ("5–15 candidate head terms", "10–20 neutral prompts"), and an exact report structure with table columns, link URL and link text. | 5 / 5 |
Workflow Clarity | Five clearly sequenced steps with explicit checkpoints and error-recovery loops: reuse research "under 30 days old" instead of paying again, "retry once with a shorter or broader form before dropping it", "If the call fails, such as no paid plan, skip it and say so", ask once before spending credit on `explore_prompt`, and fall back to chat delivery if the `seo-report` skill is unavailable. These are genuine feedback loops, matching the 5 anchor. | 5 / 5 |
Progressive Disclosure | Well-organized sections (Goal, Project context, Tools, Workflow, Deliverable, Guardrails) with all content operationally necessary inline and no bundle files to reference. At ~95 lines the skill exceeds the size where the rubric grants a no-references 5, and the detailed report-format spec ("What to return") is a borderline candidate for a separate reference file, so this sits at the 4 anchor's 'good structure, minor organization gaps' rather than 5. | 4 / 5 |
Total | 19 / 20 Passed |