github-issue-triage

Automate GitHub issue triage by launching a separate background task for each issue and streaming results in real time. This parallel processing approach handles large repositories by exhaustively collecting issues and surfacing urgent ones immediately as each task completes.

Real-time streaming works by polling background_output() for each individual issue task. As soon as a specific issue analysis completes, its results are emitted immediately rather than waiting for all issues to finish, allowing maintainers to act on urgent findings instantly.

You can process hundreds of GitHub issues without blocking by running each issue as an isolated background task with run_in_background=true. This allows your main workflow to continue operating while issues are analyzed in parallel and results are streamed as each task completes.

The best way to surface urgent GitHub issues is using a 1:1 task per issue model that streams results as each analysis finishes. This ensures critical issues are reported immediately during exhaustive pagination, rather than waiting for the entire backlog to process.

You need background task support to run real-time GitHub issue analysis, specifically the ability to run tasks with run_in_background=true and poll background_output(). Without this environment capability, the parallel streaming and per-issue isolation cannot function.

GitHub issue triage requires exhaustive pagination to collect all issues before launching background tasks. If it stops early, the process likely did not complete full pagination, preventing the subsequent parallel task launch and real-time streaming from initiating across the remaining backlog.