nemoclaw-maintainer-triage

Batch triaging GitHub issues and pull requests uses an AI workflow that fetches up to 50 unlabeled open items, generates label suggestions from a live instruction guide, and applies changes via GitHub CLI only after user approval. It logs each session to maintain triage consistency.

Batch labeling unlabeled GitHub issues is handled by iterating through up to 50 open items, generating labels and triage comments based on a live instruction guide, and applying GitHub CLI label edits only after explicit user approval to maintain consistency.

Yes, AI-assisted GitHub label triage requires explicit user approval before applying changes. The workflow generates suggestions and triage comments, waits for user confirmation, then uses GitHub CLI to apply labels and posts comments while logging each decision.

Keeping AI-generated triage comments aligned with the latest GitHub labeling instructions is achieved by loading triage-instructions.md at runtime, ensuring the AI generates labels and comments that strictly follow the most current formatting rules and guidance.

Yes, you can triage a single GitHub pull request by specifying its number. The workflow supports single-item processing to generate labels and triage comments, apply changes via GitHub CLI after approval, and log the session alongside batch capabilities for up to 50 items.

The approval-based workflow limits throughput by requiring manual confirmation for each item before applying labels via GitHub CLI. It processes up to 50 unlabeled open items per session and relies on runtime loading of triage-instructions.md to generate accurate label suggestions.