GitHub Launches Agent HQ to Manage Multiple AI Coding Agents

GitHub has announced Agent HQ, a new framework that allows enterprises to manage multiple AI coding agents within a single platform. The system integrates agents from providers including Anthropic, OpenAI, Google, Cognition, and xAI directly into the GitHub environment.

Agent HQ functions as a unified control plane that gives developers centralized oversight of various AI coding tools. Through a feature called Mission Control, users can assign and monitor agent tasks across GitHub’s web interface, VS Code, mobile apps, and the command line. The platform includes granular permission controls, allowing agents to operate within defined security perimeters and commit only to authorized branches.

New features introduced with Agent HQ include AGENTS.md files, which let teams define custom agent behaviors and enforce coding standards, and Plan Mode in VS Code, which enables developers to collaborate with Copilot on project planning before coding begins. The GitHub MCP Registry also supports integration with third-party tools through the Model Context Protocol, allowing developers to discover and connect specialized MCP servers.

GitHub has also expanded its code quality and governance tools. Agent HQ integrates with the company’s CodeQL engine for automated code review and adds a Copilot metrics dashboard to track usage across organizations. All features are available through existing paid Copilot subscriptions, providing enterprises with a unified way to manage AI-assisted development workflows while maintaining flexibility across different agent providers.