Use Copilot CLI subagents for parallel terminal work

A commenter used GitHub Copilot CLI to split repo work across multiple subagents, run commands in parallel, and review the results live.

Done with GitHub Copilot· Advanced

What they did
They opened several GitHub Copilot CLI sessions at once, usually with GPT-5.4 set to xhigh reasoning and often in autopilot mode. Their process was to begin with /plan, then use /fleet to launch parallel subagents and /delegate for asynchronous cloud tasks, before finishing with /review and a manual check of the diffs. They also kept an eye on progress through the connected VS Code chat pane.
How it went
They said this let them hand off work they previously never had time to tackle and move faster between iterations. Even so, they still had to inspect the output and guide the agents, and one commenter warned that parallel agents can step on each other in the same repo, with worktrees offered as a fix.
Tips
This works best when someone experienced stays in the loop and checks progress often; the original user was looking in every 10–15 minutes. Before starting, make the agent pause for approval on anything that would commit, push, or otherwise change shared state.

Try it yourself with GitHub Copilot

I want to use GitHub Copilot CLI on [repo] to handle [task]. First, break the work into separate pieces that can run independently. Pause and show me the plan before you start. Then use subagents or /fleet to run those pieces in parallel, and use /delegate only for background work that fits that setup. Have each agent run the needed terminal commands and tests, then come back with a combined summary of changes, failures, and open questions. Stop and ask before any commit, push, delete, submit, or other irreversible action.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Read the original on reddit.com ↗

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.