GitHub has expanded Agents HQ, enabling AI coding agents such as GitHub Copilot, Claude by Anthropic, and OpenAI Codex to execute development tasks directly within GitHub and developer editors while preserving repository context, session history, and review workflows.
Copilot Pro+ and Copilot Enterprise developers can start agent sessions from GitHub, GitHub Mobile, and Visual Studio Code, with support for Copilot CLI expected soon.
“We’re bringing Claude into GitHub to meet developers where they are. With Agent HQ, Claude can commit code and comment on pull requests, enabling teams to iterate and ship faster and with more confidence. Our goal is to give developers the reasoning power they need, right where they need it,” Katelyn Lesse, Head of Platform, Anthropic, explained.
Starting and managing agent sessions
Claude and Codex must be enabled in repository settings before starting agent sessions. Once configured, contributors can submit a request within a repository and select a preferred agent or multiple agents to execute the task.
Agents operate within structured sessions, allowing teams to monitor progress, review completed work, and compare outputs generated by different models. This approach enables organizations to identify potential issues earlier by assigning agents to distinct review or development tasks.
Parallel agent workflows inside repositories
Making coding agent available in workflow, makes all work attached to the repository itself. Running agents in parallel, you get competing approaches an edge cases before code hardens. Context stays attached to the work instead of starting from stateless prompts. You can review agent-generated changes the same way you would review a teammate’s work.
Enterprise governance and quality oversight
Agents HQ introduces administrative controls designed to support enterprise adoption of AI-assisted development. Enterprise administrators can define which agents and models are authorized across the organization, helping enforce compliance and security policies.
GitHub Code Quality evaluates how code changes affect maintainability and reliability, providing insight into long-term code health during review workflows. Meanwhile, the Copilot metrics dashboard allows organizations to track agent-assisted development activity and adoption across teams.
Copilot code review capabilities can also flag potential issues early in the development process, helping developers focus on higher-level validation and design considerations.
Read more: