GitHub · GH-600
Success criteria and evaluation signals
GH-600 · Lesson 10 of 16 · 1:23:20
YouTube loads only after you press play. This chapter opens at its timestamp in the full course. Watch on YouTube ↗
In this lesson
Prepare for the GH-600 exam (GitHub Certified: Agentic AI Developer). In this lesson: expected outcomes and operational constraints, quantitative and qualitative evaluation signals, aligning criteria with intent beyond a green build, and signals from automated scanning: code scanning with CodeQL or SARIF, dependency review, secret scanning and push protection, rulesets that make them blocking, and the cloud agent's own validation. Includes three original practice questions.
Preview the topics
- Contoso wants to know whether its agents’ pull requests actually last, and not just whether they pass. Which quantitative signal fits?
- Contoso’s agent fixed a failing build by wrapping the failing code in a catch block that ignores every error, and the checks are green. What should the evaluation have required?
- Contoso wants to block any agent pull request that introduces a high-severity code scanning alert. What should the team configure?