Skip to content
important AI agents

GitHub Copilot offers cloud agent sessions directly in Microsoft Teams

clearly official source

GitHub Copilot can now be launched as a cloud agent session directly from Microsoft Teams by mentioning @GitHub. A dedicated code channel is created for tracking progress, and administrators can require additional approval for pull requests. Public preview, paid plans, consumes AI credits.

GitHub has made it possible to start GitHub Copilot agent sessions directly from Microsoft Teams. Simply mention @GitHub in a channel, thread or direct message to start a cloud agent session that everyone in the conversation can follow – participants can ask questions, add context and help with planning, while anyone with write permissions for the repository can initiate changes to it.

The agent creates a dedicated code channel where users can track the progress of the work and continue to guide it. Work takes place asynchronously in a secure cloud sandbox; users can continue working with the results in the terminal, the GitHub Copilot app or their preferred IDE. According to GitHub, a typical use case is handing a task to the agent during a team meeting, for example when an action item comes up during a stand-up.

The feature is in public preview and is available on paid GitHub Copilot plans. Starting an agent session from Microsoft Teams consumes AI credits, with consumption managed through a usage-based billing budget for organizations; cloud sandbox usage is billed separately and can be managed through a separate budget at the product or SKU level.

Repository administrators can now require an additional approval for every pull request attributed to the Copilot integration identity in Microsoft Teams. If a repository requires two approvals, enabling this option means pull requests created by Copilot will require three approvals – according to GitHub, this keeps a human involved in the process before work created by the agent reaches production.

What changed

Why it matters

Teams that handle tasks in Microsoft Teams can hand action items to an agent without switching tools and track how they are being addressed in a shared channel directly with colleagues. The feature also consumes AI credits and offers the option to enforce additional approval for pull requests, giving organizations a way to maintain cost control and human oversight of code created by the agent.

Two audiences, two different impacts

What this means

01

For individuals

Developers can hand a task that arises during a discussion in Microsoft Teams (for example, during a stand-up) directly to the GitHub Copilot agent by mentioning @GitHub, track progress in an automatically created dedicated code channel and then continue in the terminal, the GitHub Copilot app or an IDE.

What to do Try starting an agent session by mentioning @GitHub in a Microsoft Teams channel, thread or direct message, provided the user has a paid GitHub Copilot plan and write permissions for the repository.
More practical updates →
02

For a business

Organizations can incorporate agent work from Microsoft Teams into their development workflow, but they must manage AI credits consumption through usage-based billing budgets and can configure an additional layer of pull request approval (when two approvals are required, pull requests from the Copilot integration will require three) to maintain human oversight of code created by the agent.

Development
What to decide Set or review the usage-based billing budget for AI credits and consider requiring additional approval for pull requests created by the GitHub Copilot integration in Microsoft Teams.
More business impacts →
AI agent Automation cloud agent GitHub Copilot Microsoft Teams pull request

Check the original

Event sources

clearly official source · 1 publisher, 0 independent. We count feeds from the same owner only once.

1
GitHub Changelog (Copilot and AI features) primary source · first detected Shared agentic work with GitHub Copilot in Microsoft Teams