Run a coordinated multi-agent workflow in Copilot
A developer experiments with multi-agent setups in GitHub Copilot inside VS Code as part of their workflow.
Done withGitHub Copilot
- What they did
- The developer built a spec-driven, artifact-based agent team inside GitHub Copilot in VS Code, with stages for spec, architecture, planning, implementation, review, QA, security, integration, and docs. An Orchestrator agent ran the flow, enforced gates between stages, and relied on files like spec.md, acceptance.json, tasks.yaml, and status.json so sub-agents could reread state and continue across chats when context degraded.
- How it went
- They say it worked surprisingly well for larger features and small greenfield projects, gave more consistent results than single prompts, and asked useful clarification questions. It was slower end to end and had too much overhead for tiny quick fixes.
- Worth knowing
- The main catch was context loss in long Copilot sessions, so saving progress to artifacts on disk was central to making the workflow usable.
Try it yourself with GitHub Copilot
In VS Code, set up a multi-agent workflow for [task, e.g. adding a feature to my app] with separate roles: one agent to plan, one to implement, one to write tests, and one to review. Define each role's instructions in files in my repo and tell me how to run them in order. You're done when I can run the whole flow once on [small example task].
Want help getting this working? Join the Vibe Coding Academy ↗ Run by Alex Finn, who also runs this site.
Source: Reddit · Undated
Related use cases
Run a solo company with eight agent departments
A founder set up CEO, CFO, COO, lawyer, accountant, marketing, CTO, and improver agents that share memory and consult each other.
Assign GitHub issues to Copilot coding agent
A developer assigned a GitHub issue to Copilot and watched it work, but it started from the default branch instead of the intended feature branch.
Run coding agents from one front-end across machines and VPS
A developer uses T3 Code as a front-end for Claude Code, Codex, and pi on personal machines and a VPS, and Copilot at work, for implementing tasks and drafting commit messages and PR summaries.
Build a C compiler with a team of parallel agents
Anthropic's engineering write-up on tasking Opus 4.6 agent teams with building a C compiler.
Run agents that work Jira tickets, review PRs and send updates
Luke runs several Hermes agents that read Jira tickets, write code, review PRs, track costs, summarize docs and send updates.
Visualize your codebase and coding agents in a 3D workspace
A solo dev built a local-first 3D environment where existing coding agents are embodied as avatars, and the repo is shown as a city to browse code and run apps.
Get 5 like this every morning
The best things people got an AI agent to do, each with the prompt to try it.