Old articles rarely lose traffic all at once — they slowly drift down as facts age and competitors update their own pages. Without impression data it's hard to know which parts of the article are actually worth strengthening first, and it's easy to rewrite a section that already works while missing the one actually holding the page back.
What it does
Given an article URL, this skill reads the full text for outdated parts (old dates, numbers, recommendations that no longer apply), then checks Yandex Webmaster data for that URL — which queries it shows for, at what positions and CTR — with a focus on queries where the page is "almost" ranking. It then delivers a full rewritten version that strengthens those topics, adds missing sections, and keeps the working structure and intent intact.
What you get
The updated article as a library document, with a 5-7 point change list at the top, plus a chat digest of what was outdated, which queries were strengthened (with current Webmaster positions), and what was added. Useful for pages older than a year or two that are quietly losing ground without an obvious technical cause.