Updates GitHub GitHub Copilot Coding Agent Now Generally Available on All Paid Plans
GitHub's autonomous coding agent is now GA — assign issues to Copilot and it delivers draft pull requests. Multi-agent AI coding also enabled.
Copilot’s Autonomous Coding Agent Reaches General Availability
GitHub has made its Copilot coding agent generally available across all paid plans — Pro, Pro+, Business, and Enterprise. The feature, which lets developers assign GitHub issues directly to Copilot for autonomous implementation, is now production-ready after months in preview.
The coding agent operates entirely on GitHub’s cloud infrastructure. When you assign an issue to Copilot, it creates a dedicated branch, analyzes the codebase, writes the implementation across multiple files, runs tests, and opens a draft pull request for your review. It builds repository-specific memory over time, learning your project’s conventions and reducing the need for repeated explanations.
Multi-Agent AI Coding Arrives
In a related announcement on February 5, GitHub enabled what it calls “Agents HQ” — a system where multiple AI coding agents can execute tasks within the same repository. GitHub Copilot, Claude by Anthropic, and OpenAI Codex agents can now work on separate issues simultaneously, each operating in isolated environments.
This marks a significant shift in how development teams think about AI-assisted coding. Instead of a single AI assistant, teams can now delegate entire categories of work — bug fixes, documentation updates, test writing, dependency upgrades — to different agents running in parallel.
Code Referencing for License Transparency
On February 18, GitHub added code referencing to the coding agent. When AI-generated code matches existing public repositories, Copilot now flags it with a link to the source repository and its applicable license. This addresses one of the biggest concerns enterprise customers have about AI-generated code — understanding its provenance and licensing implications.
The feature was already available in Copilot’s IDE completions but extending it to the autonomous coding agent ensures transparency even when Copilot works without direct developer oversight.
New Copilot CLI and Spaces
February also brought major updates to Copilot’s terminal experience. The Copilot CLI now supports specialized agents, parallel task execution, and autopilot mode (Shift+Tab) for uninterrupted multi-step workflows. Combined with the new Copilot Spaces — curated context workspaces that can span multiple repositories — developers have more control than ever over what context Copilot uses when generating code.
What This Means for Developers
The coding agent GA represents GitHub’s bet that autonomous AI development is ready for production use. With the agent handling routine implementation, code review catching AI-generated issues, and code referencing ensuring license compliance, GitHub is building a complete pipeline from issue to reviewed pull request.
The free tier still includes 2,000 completions and 50 chat messages per month, but the coding agent requires a paid plan starting at $10/month for Pro.