What problem does it solve? Incoming bug reports and enhancement requests often arrive under-specified, unreproduced, or mislabeled, forcing maintainers to spend time investigating before any work can begin. This Skill investigates a single issue, attempts reproduction, and prepares a classification recommendation so the maintainer only has to confirm a decision. ## Core Features & Use Cases - Issue verification and reproduction: Reads the full issue thread, explores the affected code, and attempts to reproduce reported bugs before recommending a label. - Label state machine: Maps the project's actual label vocabulary onto a model of two category labels (bug, enhancement) and five state labels (needs-triage, needs-info, ready-for-agent, ready-for-human, wontfix). - Agent-ready briefs: Produces structured briefs with problem statement, reproduction steps, acceptance criteria, likely files, and out-of-scope items for downstream agents or humans. - Use Case: A maintainer faces a backlog of untriaged GitHub issues. They invoke the Skill on the queue, review each recommendation, and confirm labels, leaving every issue either ready for work, awaiting reporter info, or durably closed. ## Quick Start Triage issue #42 and recommend how it should be classified.