Editorial Standards
How we write prompts and guides, how we produce worked examples, and how to report a mistake.
Hand-written prompts
Every prompt in the library is written by our editors, not generated in bulk. Each follows the same structure (role, context, task, constraints, output format) and includes an instruction not to invent facts. We publish a prompt only when it serves a clear, practical purpose for building or writing a website.
Worked examples
Each topic page includes a worked example. We run the prompt with a clearly fictional business in an AI assistant (Claude, by Anthropic) and show the response, sometimes lightly trimmed for length. We label these examples. AI responses vary each time, so the examples show the kind of result to expect rather than a guaranteed output.
Accurate guides
Statements in our guides about search engines, accessibility and platform rules are based on the publishers' own documentation, such as Google Search Central and the W3C's Web Content Accessibility Guidelines. Rules change, so each page shows when it was last updated, and we review pages when guidance changes.
No fake proof
We do not publish invented testimonials, user numbers or success statistics, and our prompts tell AI assistants not to produce them either.
Not professional advice
Where topics touch on law, health or finance, we note that the content is general information and recommend review by a qualified professional.
Corrections
Found an error or a prompt that does not work well? Email contact@aiwebsiteprompts.ai. We review every report.