lodging-vetlisted

# lodging-vet Choose a stay the way an experienced traveller does: **not by the star average, but by the recurring defects that average conceals** — and only the ones that matter for *this* party. A 4.2★ hotel with no lift is unremarkable for one traveller and disqualifying for a companion who can't climb stairs. A 4.6★ with self check-in by code is fine at 3pm and a locked door at midnight. The rating cannot tell you which; the reviews can, if you read them for structure instead of sentiment. ## Retrieval: what this skill can and cannot do **It cannot promise to read a listing URL.** Major booking platforms render listings in JavaScript — a plain fetch of an Airbnb listing returns a near-empty shell with no review content. Whether a URL works depends entirely on whether *your* agent has a real rendering browser; most fetch tools do not. So: 1. **Take URLs anyway** and attempt retrieval if the tool can browse. 2. **If retrieval fails or returns no reviews, say so in one line** — never imply you read a page you didn't — and ask for the fastest substitute: *paste the review text (the critical ones especially), or the property's own description of access and check-in.* 3. **Vet whatever you have.** The rubric works identically on fetched text, pasted reviews, or the user's own summary. Retrieval is a convenience; the rubric is the skill. Never infer a property's characteristics from its name, brand, price, or star count. ## How to respond 1. **Restate the constraints yo