Red Teaming
Agents
Create and manage red team agents—the AI systems you adversarially test in the dashboard.
Agents (/app/red-team/agents) are the AI systems you configure for adversarial testing. Each agent represents a target with its own test history and results.
What you can do
- List agents available to your organization
- Open an agent detail page to view configuration and test runs
- Start new adversarial test campaigns from the agent view
Setup guidance
- Name agents clearly (e.g. by product, environment, or model stack)
- Document which production or staging system each agent maps to
- Pair agent tests with the repositories that supply their tools or context
Detail pages
Selecting an agent opens /app/red-team/agents/[id] where you can review past runs, findings, and status.