Step-by-step skill execution referencing cognitive phases:
0 SETUP — Download Resources
Clone or download scripts from `https://github.com/THIAGONOMA/geo-seo-skills`
Place them in `skillschain/resources/geo-seo/`
Install Python dependencies: `pip install -r skillschain/resources/geo-seo/requirements.txt`
Verify scripts are available: `ls skillschain/resources/geo-seo/scripts/`
- Fetch homepage HTML via mcp-browser or WebFetch Extract key pages from sitemap.xml or internal links (max 50 pages)
Read and parse robots.txt for AI crawler directives
Detect business type by analyzing homepage patterns
2 CONTEXTUALIZE — Parallel Analysis
Run resource scripts for data collection:
`python skillschain/resources/geo-seo/scripts/fetch_page.py <url> full`
`python skillschain/resources/geo-seo/scripts/citability_scorer.py <url>`
`python skillschain/resources/geo-seo/scripts/brand_scanner.py "<brand>" <domain>`
`python skillschain/resources/geo-seo/scripts/llmstxt_generator.py <url> validate`
Analyze outputs across 5 tracks: AI Visibility, Platform, Technical, Content, Schema
3 HYPOTHESIZE — Score each content block
Use `citability_scorer.py` output to evaluate passages (134-167 word optimal blocks)
Use `fetch_page.py robots` output to check 14+ AI crawlers
Use `brand_scanner.py` output to map brand presence across 10+ platforms
4 EVALUATE — Composite scoring
Calculate weighted GEO Score (0-100)
Classify all findings by severity and impact
Prioritize by effort-to-impact ratio
5 RECOMMEND — Generate deliverables
Produce structured GEO audit report
Generate missing JSON-LD schemas
Create or improve llms.txt
Build prioritized action plan
Write `GEO-AUDIT-REPORT.md`
Write `llms.txt` if needed
Write JSON-LD snippets for implementation
Optionally generate PDF report with charts
7 REFLECT — Validation and telemetry
Validate generated assets
- Report execution telemetry via mcp-skillschainSuggest monitoring cadence