Content
82%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 body: exact commands, concrete extraction patterns, and a useful troubleshooting table. Remaining improvements are minor — a duplicated overview section, no output-verification step, and non-portable hardcoded template paths.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Largely lean — terse prerequisites ('No Python dependencies. No ReportLab. No JSON data wrangling.'), a compact field-extraction table, and a troubleshooting table — but 'How It Works' overlaps somewhat with 'Workflow' and could be trimmed. | 4 / 5 |
Actionability | Fully executable, copy-paste-ready pandoc and Chrome headless commands with exact field-extraction patterns (e.g. '## Overall GEO Score: XX / 100', '**Domain:**') and an explicit fallback rule ('If a field is not found... omit that --metadata flag') covering the common cases. | 5 / 5 |
Workflow Clarity | Five clearly sequenced steps with a prerequisite check (Step 1) and a troubleshooting table for error recovery; the minor gap is no explicit verification that GEO-REPORT.pdf was created and is non-empty before reporting completion in Step 5. | 4 / 5 |
Progressive Disclosure | Well-organized sections with styling and JavaScript logic appropriately externalized to the template and CSS files, clearly signaled with full paths and a 'Customizing the Report' section. Minor gaps: references use hardcoded absolute paths (~/.claude/skills/geo/templates/) rather than skill-relative paths, and no bundle files ship alongside SKILL.md. | 4 / 5 |
Total | 17 / 20 Passed |