Writing meta tags that match search intent, stand out among similar snippets, and avoid sounding machine-generated is one of the more repetitive tasks in SEO work — one that usually ends in a guessed compromise, without knowing which variant would actually get more clicks on that specific page.
What it does
Given a page URL (and optionally a target query), this skill reads the page's content, works out its topic and intent, and writes three title+description pairs: query-led (closer to the exact query), click-led (numbers, specifics, a reason to click), and a balanced option most people end up choosing. Each pair comes with an exact character count so you can see what fits the snippet.
What you get
A chat reply, no separate document: a one-line diagnosis of the current meta tags, then the three pairs with lengths and a note on when each fits best — ready to paste straight into a CMS. Handy both for a single page before publishing and for a bulk pass over pages with low click-through rates.