triaging-issues

Automating GitHub issue triage involves routing new issues to oncall teams, applying validated labels, and posting canned responses to reduce manual sorting effort. This skill uses predefined templates and scripts to classify issues as questions, bugs, or feature requests.

Routing GitHub issues to oncall queues automatically is done by validating labels against an allowlist and redirecting issues to the appropriate team. This skill applies hooks before and after triage actions to ensure consistent routing and annotation.

GitHub issue triage works with predefined labels and response templates by validating requested labels against a central allowlist in labels.json and using templates.json for standardized responses, ensuring consistent classification and communication.

You need the gh CLI to automate issue classification and labeling because this skill depends on it to interact with GitHub repositories, validate labels, and execute triage scripts for routing and annotating issues.

You can prune invalid or forbidden labels during GitHub issue triage by validating requested labels against a central allowlist. The skill automatically removes labels that do not match the predefined configuration in labels.json.

Limitations of using scripts for GitHub issue triage include reliance on static label allowlists and template configurations, meaning the system targets specific workflows like PyTorch repositories and may require manual updates for new label schemas.