Content
85%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, actionable skill body with a clear multi-step workflow, explicit validation feedback loops, and proper progressive disclosure to a real reference file. The main weakness is repeated fabrication messaging and template-with-placeholder output formats rather than fully worked examples.
Suggestions
Consolidate the fabrication/no-fabrication guidance into a single authoritative section and reference it from Step 5 and the checklist instead of restating the rationale each time.
Add one fully worked audit and one fully worked optimize example with literal filled-in values (a real URL, a real GEO Score, a real rewritten snippet) so the output templates are unambiguous.
Tighten the 'Coordination With Other Skills' and 'Critical: No Fabrication' sections, which overlap slightly with material already covered in Step 5's quality gate.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense and mostly assumes Claude's competence (no explanations of what AI search engines, robots.txt, or JSON-LD are), but fabrication warnings are repeated across the 'Critical' section, Priority 2, Step 5, and the pre-publish checklist, which could be consolidated. | 4 / 5 |
Actionability | Provides concrete executable guidance — a copy-paste bash find snippet, exact density targets (≥5 numbers, ≥1 citation/500 words, ≥3 source types), named crawlers, and JSON-LD field lists — but the audit/optimize output formats are templates with placeholders rather than fully filled worked examples. | 4 / 5 |
Workflow Clarity | Clear six-step sequence with an explicit validation gate in Step 5 (named checks plus a fabrication hard-fail with a 'don't deliver / find real evidence' feedback loop) and a pre-publish checklist, appropriate for a content-rewriting skill. | 5 / 5 |
Progressive Disclosure | SKILL.md is an overview that signals a single one-level-deep reference (references/geo-techniques.md, which exists at 592 lines) and states exactly what lives there; no nested references, and the sibling-skill coordination section points outward cleanly. | 5 / 5 |
Total | 18 / 20 Passed |