Agent skill · magnus919
aeo
Use when designing, implementing, or measuring Answer Engine Optimization (AEO) for AI answers, citations, generative search, or agent-readable web content. Do not use this as a general SEO audit or CMS operations guide; route those parts to seo-audit or the relevant platform skill.
What it needs
About 4k tokens when loaded.
What this skill does
Answer Engine Optimization (AEO) A source-disciplined implementation methodology for making useful, attributable answers discoverable and reusable by AI search and answer systems without treating vendor folklore as a ranking guarantee. When to use Load this skill when the work involves: planning or implementing AEO, GEO, LLMO, AI-search visibility, or citation visibility; turning a topic into question clusters, answer-first pages, evidence blocks, or reusable knowledge assets; implementing or validating answer-oriented HTML, JSON-LD, llms.txt, crawler controls, sitemaps, freshness signals, or Markdown delivery; designing prompt sets, citation observation logs, share-of-voice experiments, or AI-answer measurement; translating an AEO audit into a bounded implementation plan. Do not load this skill alone for broad technical SEO, keyword research, page-speed work, CMS administration, or ordinary copy-editing. Route those concerns to seo-audit, a platform skill, or the appropriate writing skill. AEO cannot guarantee inclusion, ranking, citation, or traffic. When not to use Do not use this skill alone for broad technical SEO, keyword research, page-speed work, CMS administration, or ordinary copy-editing. Route those concerns to seo-audit, a platform skill, or the appropriate writing skill. Operating model AEO is an implementation loop, not a bag of hacks: 1. Scope the audience, entities, questions, answer surfaces, business outcome, and content boundary. 2. Research primary platform guidance and the subject's authoritative evidence. 3. Map question clusters to pages and record the intended answer, evidence, entity, freshness, and owner. 4. Implement human-useful content first, then machine-readable structure and discovery controls that the target systems actually support. 5. Measure with a frozen prompt set and platform-native data where available. Record the full answer, citations, URL, date, model/surface, and failure mode. 6. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills magnus919/aeo