What problem does it solve?
It prevents teams from wasting time on stale Jira work by using historical delivery velocity to identify which backlog items are statistically unlikely to be completed and should be considered for closure.
Core Features & Use Cases
- Velocity-based scoring: Calculates team velocity and cycle-time distributions from the last six months, then assigns each backlog item a 0–100 closure likelihood score.
- Safety-first recommendations: Protects active work (e.g., In Progress / QA / Review) from being recommended for closure, while focusing on stale, blocked, past-due, or high-confidence items.
- Jira-ready outputs: Generates a scored CSV with hyperlinkable tickets and Markdown reports, and can optionally create two Jira filters for bulk operations—one for “Close Now” and one for “CVEs/Discuss.”
- Use case example: In a quarterly backlog hygiene session, run the skill for your team’s Jira JQL, review the “Close Now” recommendations in the generated CSV/report, and then bulk-close the recommended items using the created Jira filters.
Quick Start
Run the backlog purge analysis for your team by invoking /backlog-purge in Claude Code and providing your Team JQL query, Jira base URL, email, and Jira PAT when prompted.
