GitHub Agent HQ Now Supports Claude and Codex: The Future of Multi-Agent Coding Is Here

GitHub Agent HQ Now Supports Claude and Codex: The Future of Multi-Agent Coding Is Here

GitHub has significantly expanded its AI coding agent capabilities by adding support for Anthropic's Claude and OpenAI's Codex through Agent HQ. The feature is now available in public preview, allowing Copilot Pro+ and Enterprise users to run multiple coding agents directly inside GitHub, GitHub Mobile, and Visual Studio Code.

Primary sources:

• •GitHub Blog — GPT-5.3-Codex is now generally available for GitHub Copilot (February 9, 2026)

https://github.blog/changelog/2026-02-09-gpt-5-3-codex-is-now-generally-available-for-github-copilot/

• •InfoWorld — GitHub previews support for Claude and Codex coding agents (February 6, 2026)

https://www.infoworld.com/article/4130352/github-previews-support-for-claude-and-codex-coding-agents.html

What Is GitHub Agent HQ?

Agent HQ is GitHub's unified platform for running AI coding agents. Originally launched with GitHub Copilot as the default agent, it now supports multiple third-party agents, giving developers flexibility to choose the best AI for different coding tasks.

"We're bringing Claude into GitHub to meet developers where they are," said Katelyn Lesse, head of platform at Anthropic. "With Agent HQ, Claude can commit code and comment on pull requests, enabling teams to iterate and ship faster and with more confidence."

Claude and Codex: Choosing Your Coding Partner

With Agent HQ, developers can now select from three AI agents:

• •GitHub Copilot — Microsoft's in-house solution
• •Anthropic Claude — Known for reasoning and code quality
• •OpenAI Codex — The latest GPT-5.3-Codex model, now generally available

GPT-5.3-Codex delivers up to 25% faster performance than its predecessor on agentic coding tasks, with improved reasoning and execution in complex, tool-driven workflows.

Key Features of Multi-Agent Coding

Parallel Agent Execution

Developers can run multiple agents simultaneously to compare approaches and surface edge cases before code hardens. This helps identify trade-offs early in the development process.

Context Preservation

Agents operate inside the user's repository, issues, and pull requests—maintaining context rather than starting from stateless prompts. This results in more accurate and relevant code suggestions.

Integrated Review Workflow

Agent-generated changes appear as draft pull requests and are reviewed the same way human code would be, avoiding the need for new review processes.

Availability

• •Claude and GPT-5.3-Codex are available to Copilot Pro+ and Copilot Enterprise users
• •Support includes Visual Studio Code, github.com, GitHub Mobile, and GitHub CLI
• •Copilot CLI support is coming soon

Business and Enterprise administrators must enable the GPT-5.3-Codex policy in Copilot settings.

What This Means for Developers

The addition of Claude and Codex to GitHub Agent HQ represents a shift from single-agent to multi-agent development workflows. Developers can now assign different agents to different stages of a project—using Claude for reasoning-heavy tasks and Codex for speed-optimized code generation.

This flexibility, combined with parallel execution and integrated review workflows, positions GitHub as the central hub for AI-assisted software development in 2026.

Want to run your own AI coding agent outside of GitHub? Check out our AI Agent Setup Service for OpenClaw and Hermes Agent.