Create a safe project boundary.
Set the target URL, allowed hosts, session limits, and any explicit test-account notes. Every session runs inside a bounded browser sandbox.
How it works
From monitored journeys to verified work items.
UserProof owns the browser, evidence, session limits, Card QA, kanban workflow, and rerun history. Engineering agents or your team own code changes.
Generate personas and missions.
Model realistic users with goals, objections, device assumptions, accessibility context, and success criteria.
Run browser sessions.
Playwright executes bounded browser actions and stores screenshots, URLs, visible text, console errors, and network failures.
Review QA-checked findings.
Deduped cards include severity, confidence, repro steps, expected vs actual behavior, acceptance criteria, and a Card QA result.
Prompt, fix, and rerun.
Ready cards include compact agent prompts. Verify cards require rerun evidence before they become resolved.
Product boundary
UserProof finds and verifies. Your workflow fixes.
That is the difference between a coding agent and a continuous product-testing system. UserProof keeps discovering, ranking, evidencing, deduping, and rerunning journeys after the code changes.
Discover issues
Capture evidence
QA finding cards
Verify after rerun
Accept cards
Fix customer-facing behavior
Use agent prompts
Request reruns
Ready when your next release is