GitHub Copilot Cloud Agent Is Now Generally Available for Linear, Launching Isolated Environments from Issues and Returning Draft PRs

GitHub announced on July 23 that the Copilot cloud agent for Linear is generally available, connecting previously separate work-planning and asynchronous coding-agent workflows into a single process. Users can delegate an issue to Copilot in Linear; the agent reads the issue, works independently in an ephemeral development environment provided by GitHub Actions, creates a draft pull request, and continuously posts progress to the Linear activity timeline before requesting human review when finished.

Compared with the previous integration, the generally available release adds controls with greater engineering significance. For each issue—or through Linear workspace and team guidance—teams can specify the model, a custom agent stored in the repository, the target base branch, and the agent’s working branch. While the agent is running, users can also mention Copilot in comments to add instructions or correct its direction. Linear guidance can define repository selection, how commits and issues should reference each other, and the review process, and it supports Markdown editing and version history.

The responsibility model is noteworthy: Linear treats the agent as an installable app user that can be mentioned, assigned issues, and reply to comments, while a human remains the primary assignee and accountable owner. Installing GitHub Copilot for Linear requires GitHub organization owner and Linear workspace admin permissions; administrators can also restrict which teams the agent can access. This means the integration does not move the entire codebase into Linear. Instead, the trust boundary is jointly formed by application permissions on both platforms, issue context, and the GitHub execution environment.

Before adoption, teams should assess whether Linear descriptions, comments, or guidance could introduce untrusted text into the agent prompt. They should also restrict GitHub tokens, Actions secrets, network access, and writable branches. GitHub has not published benchmarks for accuracy, PR merge rates, or cost; general availability indicates that the product interface has matured, not that the agent can reliably handle every issue. Areas to watch include the portability of custom agents, recovery after failures in long-running tasks, and whether cross-system audit logs can fully trace every code change.