issue-triage

Automate GitHub issue triage by fetching open issues and evaluating them against your codebase to verify relevance. The process analyzes referenced files, checks if the problem exists, and applies YAGNI criteria to return a structured verdict recommending whether to keep or close each issue.

Prune noisy GitHub backlogs by screening open issues through concurrent evaluation to classify obvious closes and keeps. Remaining issues undergo deep evaluation against the current codebase state, producing actionable triage results that reduce cognitive tax from irrelevant or outdated requests.

YAGNI evaluation criteria for issue management works by analyzing referenced files in your codebase to verify if a reported problem actually exists. It then applies the "You Aren't Gonna Need It" principle to decide whether to keep an issue open or close it with standard traceability comments.

You can merge duplicate GitHub issues during automated triage. After evaluating issues against your codebase and classifying them, the system presents triage results and planned actions that include closing, keeping, or merging duplicates before executing approved actions with standard close comments.

Issue triage works with both specific scopes and all open GitHub issues. Stage one determines the scope from user input or fetches all open issues, allowing you to target a specific subset of your backlog or evaluate the entire repository queue concurrently.

Evaluating GitHub issues against real code requires accessible file references within the codebase to verify if the problem exists. Issues lacking clear file references or describing unbuilt features may face limitations during the deep evaluation stage when applying the YAGNI keep or close criteria.