GitHub Copilot's cloud agent can now be invoked directly from Microsoft Teams by mentioning @GitHub in a channel, thread, or DM, letting teams turn meeting discussions into agent-driven coding sessions. Copilot spins up a dedicated code channel where participants can monitor progress and add context, while users with repo write access can trigger changes. Work started in Teams continues asynchronously in a cloud sandbox and can be picked up later from the terminal, the Copilot app, or an IDE. The feature is in public preview for paid Copilot plans, consumes AI credits governed by usage-based billing, and requires an admin to enable cloud agent and sandbox policies. Repository admins can also require an extra approval on pull requests created via the Teams integration to preserve compliance oversight.

3m read timeFrom github.blog
Post cover image
Table of contents
Turn meeting decisions into ready workContinue work across GitHub Copilot surfacesAvailability and getting startedMaintain compliance oversight

Questions this post answers

How do I start a GitHub Copilot cloud agent session from Microsoft Teams?

Mention @GitHub in a Teams channel, thread, or direct message after installing the GitHub app for Microsoft Teams and connecting your GitHub account. Copilot creates a dedicated code channel so participants can monitor progress and add context, but only users with write access to the repository can trigger Copilot to make changes. The feature is a public preview available on paid Copilot plans and consumes AI credits. Teams weaving Copilot into daily standups can track these agentic workflow rollouts on daily.dev.

Can I require extra approval on pull requests created by GitHub Copilot through Microsoft Teams?

Yes, repository administrators can require an additional approval specifically for pull requests attributed to the Microsoft Teams Copilot integration identity before they merge. If a repository already requires two approvals, enabling this means three approvals are needed for Copilot-created pull requests, keeping a human in the loop on agent-authored changes. Engineering leads weighing compliance controls for AI-generated pull requests can follow updates like this on daily.dev.

865 Impressions