What this prompt does
This prompt builds a Generative Engine Optimization (GEO) strategy so your content gets cited inside AI-generated answers rather than buried beneath them. You supply [site_url] and [topic_area], and the AI audits and plans across ten fronts: passage-level citability, an llms.txt implementation, an AI-crawler accessibility check (GPTBot, ClaudeBot, PerplexityBot in robots.txt), brand-mention signals, AI-parseable structured data, content-format optimization, citation-worthy snippets for [target_queries], E-E-A-T signals for [author_expertise], a monitoring approach via [monitoring_method], and a content calendar focused where AI Overviews are appearing.
The structure works because AI search rewards different things than classic SEO. Generative engines lift clear, self-contained passages — definitions, tables, direct answers — so the prompt reframes each page as a set of citable units. By pairing that with crawler accessibility and authority signals, it covers both whether AI systems can read you and whether they will trust you. The [target_queries] variable focuses snippet creation on the exact questions you want to be cited for, rather than chasing visibility in the abstract.
When to use it
- AI Overviews or ChatGPT are answering your niche's queries directly and your traffic is slipping.
- You want to audit whether GPTBot, ClaudeBot, and PerplexityBot can actually crawl your site.
- You need an llms.txt file and don't know what to put in it.
- You want to restructure content into citable passages for
[target_queries]. - You are building authority signals around a specific
[topic_area]. - You need a way to monitor whether you appear in AI answers at all.
Example output
Expect a ten-part GEO plan: a citability audit with concrete restructuring advice, a draft llms.txt, a robots.txt accessibility checklist for the named AI crawlers, recommended structured-data types, snippet drafts targeting [target_queries], E-E-A-T recommendations for [author_expertise], and a monitoring plan built around [monitoring_method]. The content calendar prioritizes topics where AI Overviews already surface.
Pro tips
- Set
[target_queries]to the specific questions you want cited for — vague queries produce vague snippets that no engine picks up. - Make
[author_expertise]accurate and verifiable; E-E-A-T signals only help when the credentials behind them are real. - Treat the llms.txt and robots.txt audit as table stakes — if the named crawlers are blocked, no amount of content optimization gets you cited.
- Structure each target page around a clean definition or direct-answer paragraph near the top; that is the unit generative engines lift most readily.
- Choose a
[monitoring_method]you will actually run regularly, since AI-answer visibility shifts faster than classic rankings. - Build brand-mention signals deliberately within your
[topic_area], because generative engines lean toward sources they already see referenced as authorities. - After the first pass, ask the model to rewrite one existing page into citable passages as a worked example you can pattern-match across the site.
