GitHub brings Copilot coding-agent sessions into Slack and Microsoft Teams

GitHub has moved Copilot's cloud coding agent closer to where many software teams already triage work: Slack and Microsoft Teams. In separate August 21 changelog posts, the company said developers can mention @GitHub in a channel, thread, direct message, or meeting chat to start an agent session that can answer repository questions, investigate problems, update issues, implement changes in a secure cloud sandbox, and open a pull request when work is ready for review.

The practical shift is less about chat convenience than accountability. Instead of one developer privately prompting an agent and later explaining the result, GitHub is making the agent session visible to the group that raised the problem. That may help teams preserve context from incident threads, planning discussions, and standups, while still keeping the final merge decision inside normal repository controls.

GitHub says the Slack preview works for organizations on GitHub Copilot Business and GitHub Copilot Enterprise, with usage counted against existing Copilot entitlements and managed through cloud-agent budgets. For Microsoft Teams, the preview is available with paid Copilot plans; GitHub says cloud-agent sessions consume AI credits, while cloud sandbox usage is billed separately and can be controlled with product-level or SKU-level budgets.

What teams should check

• Confirm whether Copilot cloud agent and cloud sandbox policies are enabled before telling developers to use the integrations.
• Decide which channels or teams should be allowed to start agent sessions, especially in repositories with regulated workflows.
• Review budget settings for AI credits and cloud sandbox usage so experimentation does not become an unplanned cost center.
• Use the extra-approval option for pull requests attributed to the Copilot app or Microsoft Teams integration identity when human signoff is required.

The social hook is simple: the next AI coding-agent experiment may start in the team chat, not in the IDE. That is useful, but it also changes the review surface. A conversation thread can become a work request, a planning document, and an audit trail, so teams should treat prompts and shared context as part of the engineering process.

One likely effect is that engineering managers will judge agent tools less by single-user autocomplete gains and more by whether they reduce handoff loss between chat, issue trackers, pull requests, and local development. It remains unclear how quickly teams will standardize guardrails for these shared sessions, but GitHub's approvals, permissions, and budget controls show that enterprise adoption now depends as much on governance as on model capability.

Sources

• github.blog
• github.blog

Cover photo by Christina Morillo on Pexels, used under the Pexels License.