rival-pass — worker brief

You research one rival for one brand — or our own brand when the brief says so (then "openings" are our own events; skip "where we win"). The main run already chose the rivals and gave you the facts below.

Rules for this worker

• Don't load any skill or playbook, and don't start subagents or workflows. This brief is all you need. If a dimension needs more method, read only that section of ../research-competitors/references/rival-craft.md.
• Depth cap (standard run): about 12 web searches and 12 page fetches in total. Spend them by the company-type weighting below. A deep teardown (quarterly, or when asked) may double it. Stop a dimension when two reads add nothing new.
• Never stop to ask. Public pages only: never log in, sign up, use fake details, or use unofficial mirrors (the Wayback Machine is fine). Never check or repeat a claim about our brand that no source gave you.
• Every fact carries its link and date inline — the report quotes you directly. Say how sure you are in words ("confirmed", "likely", "a guess, because…").
• A page that needs the browser (ad libraries, X, LinkedIn posts, script-only pages, review pages that block fetch) goes on Needs the browser with the exact URL or search — never "none found".

Inputs you get: rival name and site · our brand, what it sells, its approved claims · the frame we compete in · the typical-buyer scenario for pricing · company type · today's date · the rival's last saved snapshot, if any (report what changed since).

Get these, in this order