browse-aeo skill

Answer Engine Optimization — page audit for AI answer readiness + SERP analysis for Google AI Overviews and Perplexity.

Install

npx skills add https://github.com/ulpi-io/skills --skill browse-aeo

/browse-aeo <url-or-query> — or invoked when the user asks for:

Two modes depending on the input:

Analyze a page for AEO readiness (whether AI engines will cite it):

Navigate with browse goto <url>
Extract structured data with browse schema — FAQ, HowTo, and Q&A schemas are critical for AI citation
Extract meta tags with browse meta — description quality matters
Check heading hierarchy with browse headings — clear H1-H2-H3 = answer-friendly
Extract text with browse text — scan for FAQ patterns, direct answers, concise definitions

Scores across 5 weighted dimensions:

Dimension	Weight	What it measures
Structured Data	25%	JSON-LD/Microdata coverage, FAQ/HowTo presence
Meta Quality	15%	Description length, canonical, OG completeness
Heading Structure	20%	H1 count, Q&A patterns in H2/H3
Answer Readiness	25%	Direct answer paragraphs, definition patterns
Authority Signals	15%	Author schema, publisher, dates

Check how a query appears in AI-powered search:

Navigate to Google: browse goto "https://www.google.com/search?q=<query>"
Take interactive snapshot: browse snapshot -i
Agent reads the snapshot to identify AI Overviews, featured snippets, People Also Ask
Navigate to Perplexity: browse goto "https://www.perplexity.ai/search?q=<query>"
Analyze citations — which domains get cited, position, context

For Google/Perplexity, use camoufox to avoid bot blocks — both aggressively gate non-browser requests.

goto, snapshot, schema, meta, headings, text, wait