GitHub Copilot in Slack and Teams Is GA: What Small Teams Should Set Up First

What GitHub shipped

• @GitHub in Slack and Teams is generally available (November 6). GitHub says you can turn conversations into code: investigate bugs, create issues, implement changes, and open pull requests without leaving the conversation. In Teams, Canvas lets people review implementation plans, code changes, and previews alongside the discussion. @GitHub now identifies the repository from your message and launches a session without asking you to pick it again.
• Clearer attribution. Issues and pull requests created by @GitHub now include "Created from a Slack/Teams conversation with @username" in their description.
• Copilot code review settings through the API (November 6). Administrators can now manage code review settings through the REST and GraphQL APIs, from enterprise policy down to individual repositories. Copilot's reviews and comments can be identified through explicit attribution fields, and code review sessions now show the AI credits consumed.
• More agent usage data (November 6). The Copilot impact dashboard's agent adoption chart now includes Copilot Chat and the VS Code Agents window, for nine Copilot experiences in total. The "Passive users" cohort is now called "Chat-first users."
• Copilot in Xcode's built-in coding assistant (November 4). This is a public preview starting with Xcode 27.3 beta. The existing GitHub Copilot for Xcode extension remains a separate option.

What Microsoft changed in Copilot Studio

On October 28, 2026, Microsoft said the GitHub Copilot harness in Copilot Studio is now called the Microsoft Copilot harness, reflecting a shared foundation for agent reasoning, tool use, and orchestration across Microsoft's Copilot products. Microsoft says existing agents keep working with no changes to capabilities or billing, and the new name will appear in the product over the coming weeks.

Microsoft also published internal evaluation figures comparing it with the Standard harness. It reported higher pass rates on the same tasks (10.8 percentage points with an OpenAI model and 22.1 points with an Anthropic model) and said agents kept working after a failed tool call in 84 to 91% of evaluated cases. These are Microsoft's own internal numbers, and Microsoft notes they may vary by model and configuration. The same post points to recently announced features for persistent, proactive agents, including agent accounts, business process memory, and triggers.

Why this matters for small and mid-size teams

Until recently, starting AI coding work meant someone opening an editor or the GitHub site. Now a request in a busy Slack or Teams channel can become a coding session and a pull request. That is a real productivity gain for small teams. It also means your chat workspace becomes part of how code changes are started, so chat permissions, channel membership, and repository access matter more than before.

Agents that keep retrying after a failure are also more capable and less predictable. That is useful for long business processes, and it is a reason to keep a person in charge of approvals.

What to set up first

1. Decide who can start agent work from chat. Limit it to people who already have write access to the repositories involved, and keep it out of channels that include clients or contractors unless that is intended.
2. Keep human review in front of every merge. Use branch protection so pull requests opened by @GitHub or any agent need an approving review from a person.
3. Use the attribution. The "Created from a Slack/Teams conversation" note and the code review attribution fields make it easier to audit what came from an agent. Make that part of your review habit.
4. Set code review policy centrally. If you have several repositories, use the new settings controls so review effort and tool access are consistent instead of configured repo by repo.
5. Watch usage and cost. The impact dashboard and AI credit reporting show which features people actually use. Review them monthly before expanding licenses.
6. Inventory your Copilot Studio agents. If you build agents in Copilot Studio, note which harness each one uses and what each agent is allowed to do, especially agents that run on triggers.

How NTx can help

Turning on new AI features is easy. Deciding who can use them, what they can reach, and how work gets reviewed is the part that protects a small business. That is the focus of our AI Consulting work. Related reading: agentic AI hunting exposed secrets and AI coding-tool supply chain risk.

FAQ

Is GitHub Copilot available in Slack and Microsoft Teams?

Yes. GitHub announced on November 6, 2026 that @GitHub in Slack and Teams is generally available. You can investigate bugs, create issues, implement changes, and open pull requests from a conversation.

How can I tell if a pull request was created from Slack or Teams?

GitHub says issues and pull requests created by @GitHub include the text Created from a Slack/Teams conversation with the username of the person who started it in their description.

What is the Microsoft Copilot harness?

It is the new name, announced October 28, 2026, for what Copilot Studio previously called the GitHub Copilot harness. Microsoft describes it as a shared foundation for agent reasoning, tool use, and orchestration. Existing agents keep the same capabilities and billing.

Can administrators control Copilot code review settings centrally?

Yes. As of November 6, 2026, GitHub lets enterprise, organization, and repository administrators read and update Copilot code review settings through the REST and GraphQL APIs, within their role.

Should a small business let AI agents open pull requests?

It can be a real time saver, as long as a person reviews and approves every change before it merges, and only people who already have write access can start agent work.

Sources

1. "@GitHub in Slack and Teams is generally available and more," GitHub Changelog, November 6, 2026. https://github.blog/changelog/2026-11-06-github-copilot-in-slack-and-teams-is-generally-available
2. "Copilot code review: API management and review overview updates," GitHub Changelog, November 6, 2026. https://github.blog/changelog/2026-11-06-copilot-code-review-api-management-and-review-overview-updates
3. "Copilot impact dashboard adds chat and VS Code agent usage," GitHub Changelog, November 6, 2026. https://github.blog/changelog/2026-11-06-copilot-impact-dashboard-adds-chat-and-vs-code-agent-usage
4. "GitHub Copilot is now available in Xcode's coding assistant," GitHub Changelog, November 4, 2026. https://github.blog/changelog/2026-11-04-github-copilot-is-now-available-in-xcodes-coding-assistant
5. Microsoft Copilot Studio Team, "The GitHub Copilot harness in Copilot Studio is now the Microsoft Copilot harness," Microsoft Copilot Blog, October 28, 2026. https://www.microsoft.com/en-us/copilot/blog/copilot-studio/the-github-copilot-harness-in-copilot-studio-is-now-the-microsoft-copilot-harness/