A coding agent can open a pull request in minutes. The team still has to decide whether the change is ready to ship.

This session looks at how to review agent-generated code and where tests and security checks fall short. We’ll discuss what should block a release, including how much confidence to place in tests written by the same agent.

After deployment, we’ll cover how to spot a failing release and check that a rollback worked. Agents can help investigate problems and propose fixes, but the team needs to agree on what they can change without approval.

The session includes a PR-to-production checklist for reviewing agent-generated changes, with release criteria and recovery steps your team can agree on before deployment.