Agent reports are red-team projects that stress-test an AI system you describe in the dashboard. They help you find jailbreaks, policy gaps, and unsafe behaviors before users encounter them.

Creating an agent report

1. Open Reports and click Create report, or go to /app/reports/new.
2. Choose Agent.
3. Enter a description of the system under test (product, environment, model stack).
4. Select an agent type (chatbot, voice agent, coding agent, AI workflow, data pipeline, or other).
5. Submit to create the project and open its detail page.

Agent detail page

Each agent project lives at /app/red-team/agents/[id]. From there you can:

• Review configuration and test history
• Launch adversarial test campaigns
• Download reports when available
• Track status as runs complete

Name projects clearly so your team can map each report to a real deployment (staging vs production, product line, etc.).

Best practices

• Document which production or staging endpoint each agent represents
• Pair agent tests with repository reports when the agent depends on specific code or tools in a repo
• Re-run tests after material model or policy changes

Legacy path

/app/red-team/agents lists the same agent projects now unified under Reports.

Next steps

• Running reports
• Findings