Trace is the part of co|op that lives right in the browser. It reads what your team types into ChatGPT, Claude, or Copilot — the same instant they type it — and steps in before anything sensitive leaves. No IT ticket, no training video. Try it yourself below, exactly like you'd type into any AI tool.
That check happened the moment you hit enter — no server round-trip, no delay. Trace works the same way on a real employee's machine.
In the real extension, Trace catches this before the message ever reaches ChatGPT — not after, not as a report the next day.
Trace warns and lets people confirm on lower-risk items. On things like SSNs or card numbers, it stops the message outright.
Case notes, SSNs, and settlement figures get pasted into AI drafting tools constantly. Trace catches it before it leaves the browser.
Medical notes, DOBs, and employee IDs show up in AI-assisted emails more than most HR teams realize.
Bank accounts, routing numbers, and EINs are exactly the kind of thing that ends up in a "can you help me write this" prompt.
Policy numbers and health details are sensitive by definition — Trace flags them the moment they're typed.
This is the dashboard your team gets from day one — every AI prompt Trace catches shows up here, scored and organized automatically. No spreadsheets, no manual review. You open it, and you already know what happened, when, and how serious it was.
Every incident is scored High, Medium, or Low the moment it's caught — no digging to find what matters.
Daily incident trends and a full recent-activity feed, so you can see patterns forming, not just single events.
This view is scoped entirely to your organization — nothing here ever mixes with another company's incidents.
Start a free pilot — no credit card required.