aiwf-aeo-geo
Use for answer-engine optimization, generative-engine optimization, AI-search visibility, citation readiness, entity clarity, source-worthy content, question coverage, and measurement across answer an
SKILL.mdskills/aiwf-aeo-geo/SKILL.md
GitHubAIWF AEO And GEO
Core Rule
Improve the usefulness, clarity, originality, and verifiability of content for people and machines. Do not promise inclusion, citations, rankings, or traffic, and do not treat unsupported AI-search hacks as requirements.
Workflow
- Define the audience, decision, entity, product facts, evidence, geography, and target question set.
- Audit whether the site makes who, what, where, why, proof, limitations, dates, and authorship unambiguous.
- Find information gaps where first-party experience, measurements, comparisons, examples, or primary data can add real value.
- Build a question-to-evidence map and write direct answers supported by nearby context and accessible source links.
- Keep claims consistent across pages, profiles, documentation, datasets, and public repositories.
- Measure eligible pages, impressions, referrals, brand/entity queries, cited pages, conversions, and answer accuracy over defined windows.
- Record observations separately from causes; answer systems are volatile and attribution is incomplete.
Read references/source-register.json before claiming that a file, schema, content format, or optimization is required by a search or answer platform.
Guardrails
- Reject mass-produced answer pages, fabricated expertise, fake mentions, citation bait, hidden text, doorway pages, and copied summaries.
- Do not claim that
llms.txt, tiny content chunks, special AI rewrites, or excessive structured data guarantees visibility. - Use structured data only when it matches visible content and a supported vocabulary.
- Add
aiwf-web-seofor rendering, canonicalization, robots, sitemaps, performance, and indexability. - Add
aiwf-startup-marketing-growthfor positioning, distribution, funnel, or campaign planning.
Output
Return prioritized question clusters, entity and evidence gaps, content changes, measurement plan, unsupported tactics rejected, and the next recheck date.