Tests authored
- MT.1113AI agents should not be shared with broad access control policies.
- MT.1114AI agents should require user authentication.
- MT.1115AI agents should not have risky HTTP configurations.
- MT.1116AI agents should not send email with AI-controlled inputs.
- MT.1117Published AI agents should not be dormant.
- MT.1118AI agents should not use author (maker) authentication for connections.
- MT.1119AI agents should not have hard-coded credentials in topics.
- MT.1120AI agents should not use MCP server tools without review.
- MT.1121AI agents with generative orchestration should have custom instructions.
- MT.1122AI agents should not have orphaned ownership.