GitHub Agentic AI Developer Course Notes

A cloud agent is used when a repository task needs more than a chat answer: Copilot receives repository context, works in an isolated branch, commits proposed changes, and opens a pull request for review. GH-600 questions often separate this workflow from ordinary chat because the cloud agent produces repository artifacts rather than only suggesting code inline.

• Confusing Copilot Chat with the cloud agent; chat can explain or suggest code, while the cloud agent creates branch and pull request artifacts.

• If the question says Copilot should work from an assigned GitHub issue and open a pull request, choose the cloud agent workflow.

Tool interaction begins with deciding which external capabilities the agent may call. MCP connects Copilot to tools and data sources through servers, while Copilot SDK custom agents define which tools are available and what permissions apply during execution.

• Treating MCP as memory; MCP connects tools and resources, while memory stores context across a session or future work.

• If the question mentions approved MCP server discovery, choose MCP registry configuration.

Short-term memory is the context available inside the current agent session or conversation. Use it for details that matter only while the current task is active, such as the files already inspected, a temporary decision, or the latest reviewer instruction.

• Confusing short-term memory with external memory; short-term context is session-local, while external memory is durable across tools or sessions.

• If context only matters during the current task, choose short-term memory.

Evaluation starts by naming what success looks like before accepting agent output. Quantitative signals include test results, lint counts, changed files, and detected regressions; qualitative signals include code clarity, maintainability, security risk, and whether the pull request actually solves the requested problem.

• Accepting a pull request because it compiles without checking whether it satisfies the original acceptance criteria.

• If the question asks how to judge agent quality, look for explicit success criteria, tests, review checkpoints, and comparison against the plan.

Multi-agent coordination uses runtime delegation when one agent cannot sensibly handle every part of a task. A coordinator or SDK runtime chooses the agent profile that best matches the requested work, tool needs, repository context, and current step.

• Giving every custom agent every available tool instead of selecting a specialized profile for the work.

• If the scenario asks who handles a specialized part of a larger task, think runtime delegation or sub-agent orchestration.

Guardrails define what an agent may do without approval and what must stop for human judgment. GH-600 questions usually tie guardrails to repository patterns, autonomy levels, branch rules, and review requirements rather than to vague reminders about being careful.

• Relying on prompt instructions for security when the scenario requires enforceable controls such as branch protection, MCP allowlists, or firewall rules.

• If the clue says required reviews, status checks, or push restrictions, choose branch protection.