issue-triage

To organize a flat issue backlog, you can reorganize it into a structured dependency graph that applies duplicate detection, cluster synthesis, and safe dependency rewires to clarify priorities and identify blockers.

The best way to find duplicate issues in a project board combines mechanical duplicate checks, graph signals, and specialist reasoning to group overlapping reports and related work into identifiable clusters.

Mapping out blockers and dependencies for open issues is done by rewiring relationships safely, recommending types like blocks, tracks, parent-child, or supersedes, and asking for confirmation before applying them.

Reorganizing a backlog with fewer than fifteen open issues is not the intended use case; this dependency graph restructuring applies to boards with duplicate reports, overlapping work, and more than fifteen open issues.

Backlog triage does generate an audit trail by creating an apply.sh script that records per-change confirmations, ensuring all dependency rewires and board modifications are fully reviewable.

To capture the current state of a project board before restructuring, take a snapshot that records open issues, cycles, stale items, and graph structure prior to making any dependency changes.