hive

The Triage Issue skill is a robust automation tool designed for software maintainers to streamline GitHub repository management. It bridges the gap between reported user issues and actual codebase implementation by performing deep technical analysis. When an issue is triggered, the agent fetches the full issue metadata, extracts core technical claims, and correlates them with existing code paths, documentation, and design logic using grep and static code investigation. This skill is intended for technical leads and repository maintainers who need to maintain a high signal-to-noise ratio in their issue trackers.

• Automatically fetches and parses GitHub issue details including title, body, state, and labels via CLI integration.

• Conducts automated code investigation by scanning relevant files, functions, and logic to verify the accuracy of a reported bug or feature request.

• Categorizes issues into actionable buckets such as valid bugs, misunderstandings, fundamentally flawed requests, or duplicates, ensuring proper triage flow.

• Drafts comprehensive, high-quality technical responses that acknowledge user concerns while providing clear evidence, code references, and industry-standard justifications.

• Facilitates human-in-the-loop (HITL) workflows by presenting findings and drafted responses for final review before taking irreversible actions like closing issues.

• Supports standardized issue closing protocols, ensuring that "Not Planned" or "Duplicate" reasons are applied correctly with clear context for the reporter.

• Requires GitHub CLI (gh) authentication and appropriate repository access permissions to read, comment, and close issues.

• Designed for use in repositories that follow structured development practices where code paths and implementation details can be clearly traced.

• Operates best when the AI has access to a consistent codebase with readable architecture and documented design decisions.

• Users should ensure the bot is configured with the correct repository path and that it adheres to internal guidelines to prevent premature closure of edge-case bugs.

• The agent performs best when tasked with verifying technical claims that depend on specific function signatures, configuration settings, or logic implementation within the repository source code.