Capture
Widget records the report, URL, browser, viewport, recent clicks, console errors, and failed requests.
Critique Intake
Critique Intake is agentic bug intake for software teams. Users report what broke, Critique asks one useful follow-up, captures technical context, and turns the report into investigated work for your Dashboard, Review runs, Builder, and issue tracker.
Invoice download is broken
Widget records the report, URL, browser, viewport, recent clicks, console errors, and failed requests.
The user gets one focused follow-up question instead of a generic request for more detail.
Critique classifies the report, estimates severity, identifies evidence, and suggests an owner.
The team gets an agent-ready prompt for Builder, Review runs, GitHub issues, Linear, or the Merge Gate API loop.
Who it is for
Reduce the “what actually happened?” loop before a developer touches the issue.
Send engineering investigated packets instead of forwarding vague customer quotes.
Keep production feedback moving without building a full bug intake stack first.
Feed coding agents better tasks before they open PRs that Review runs and Change Passports can evaluate.
<script
src="https://critique.sh/intake/widget.js"
data-project-id="PROJECT_ID"
async></script>Reports land in the Dashboard as investigated packets.
Copy the generated prompt into Builder or a coding agent run, then use Review runs and Change Passports to judge the PR.
Use Merge Gate API when intake is part of an automated support-to-fix pipeline.
The user says what broke inside your app.
Critique clarifies, captures context, and assigns evidence.
The packet becomes a prompt for Builder, an agent, or a human fix.
Start with intake. Keep the rest of the loop inside Critique.