A note in this repository, Claude/github-agentic-ai-developer/README.md: study notes for the GitHub Certified: Agentic AI Developer exam (GH-600) and course GH-600T00 Developing in Agentic AI Systems.1
Coverage
Only module 2, Designing Agent Architecture and SDLC Integration, was read unit by unit; the other five modules are summarized from their unit titles and the course and exam pages.1 Wiki pages citing this source draw on module 2.
Exam facts
- GH-600: provided by Microsoft, maintained by GitHub; proctored, 120 minutes, English only, through Pearson VUE; intermediate level.1
- Retake after 24 hours following a first failure; later waits grow.1
| Domain | Weight |
|---|---|
| Prepare agent architecture and SDLC processes | 15–20% |
| Implement tool use and environment interaction | 20–25% |
| Manage memory, state, and execution | 10–15% |
| Perform evaluation, error analysis, and tuning | 15–20% |
| Orchestrate multi-agent coordination | 15–20% |
| Implement guardrails and accountability | 10–15% |
As listed in the notes.1
Takeaways
- Agents propose work through branches and pull requests; GitHub controls decide whether it is accepted.1 See Agents propose, people and policy accept.
- Autonomy should be sized to the risk of the paths a change touches.1 See Risk-based autonomy.
- A task needs defined inputs, outputs, and success criteria.1 See Delegation contract.