Cloud agents return a branch or pull request. Issue-to-pull-request agents prepare a reviewable change, and CI or scheduled agents run maintenance tasks. The boundaries overlap, because one product may support several modes. Copilot cloud agent, for example, is also a cloud agent.
So compare agents on two axes. The first is the execution environment, with its tools and permissions. The second is how much supervision each needs and how results return to the team. Those differences matter more than the category name.
Rewriting in plainer words…
This answer doesn't lend itself to a diagram - it reads best . No credits were charged.
Saved in this browser - sign in to keep your review list.
How should your speech become text?
Listening… your words appear above as you speak - tap Stop when you're done.
Recording · cr - tap Stop & transcribe when you're done.
Transcribing with AI…
Voice:
Keep going - a few more words and AI can grade it.
Interview lens
How interviewers actually use this question
Likely follow-ups
For a small bug you want to watch closely, which kind of agent would you pick, and why?An IDE or terminal agent, so you can review and steer each step. Say what you give up compared with handing a cloud agent the task.
I group coding agents by where they run and what starts them, and that gives five common categories.
IDE agents work inside the editor, like the agent features in Cursor and GitHub Copilot, and terminal agents run in a shell with repo tools, like Claude Code or Aider.
Cloud agents such as OpenAI Codex cloud and Google Jules take a handed-off task, work asynchronously in a hosted environment, and return a branch or pull request.
Issue-to-PR agents like GitHub Copilot cloud agent take an assigned issue and prepare a reviewable change.
CI or scheduled agents, like the Claude Code GitHub Action, run in the pipeline on an event or a timer, usually for maintenance tasks.
The lines blur because one product can cover several modes, so I compare agents on environment and permissions, and on supervision and how results come back.
Weak answers to avoid
Lists product names with no categoriesA string of tool names shows you read headlines, not that you understand them. Group by where the agent runs and what starts it, then give one example each.
Treats the categories as strict boxesProducts span modes. Copilot cloud agent is both an issue-to-PR agent and a cloud agent. Say the lines overlap and compare on environment and supervision.
Counts autocomplete as an agentInline completion suggests code, it doesn't act. Agents take multi-step actions with tools, like editing files and running commands, so keep suggestion tools off the list.
Skips how results come backWhere an agent runs is half the picture. Say whether you get a live edit you watch or a PR you review later, because that decides how much supervision it needs.
Interview lens
Likely follow-ups, what you can say, and the weak answers to avoid.
