recover

Resume or restart stuck pipeline tickets with worktree and board checks.

24|5|Updated Feb 28, 2026

Search Tags:#automation #devops #pipeline #git-worktree #recover #board-api #stuck-ticket

Ready to run

npx skills add https://github.com/yves-s/just-ship --skill recover-yves-s

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: recover
Source: https://github.com/yves-s/just-ship/tree/main/skills/recover
Command: npx skills add https://github.com/yves-s/just-ship --skill recover-yves-s

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Stuck-ticket recovery: automatically resumes work when code exists or restarts when the worktree is empty.

Core Features & Use Cases

Automatic resume when there is a worktree with changes; restart when no usable worktree; board status updates optional.
Handles concurrency guard, event signaling, and cleanup steps; supports both board-enabled and standalone modes.

Quick Start

Run the recover command to recover a ticket like /recover T-501 and let it decide between resume or restart.

Frequently Asked Questions about recover

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I automate recovery of stuck pipeline tickets in a devops workflow?▼

Automating stuck pipeline ticket recovery involves checking for an active ticket and resuming the existing git worktree with changes, or restarting if the worktree is empty. This handles concurrency guards and cleanup automatically.

When do I need to resume versus restart a worktree for a stuck ticket?▼

You resume a worktree when changes already exist, but restart when no usable worktree is present. The recovery process applies status checks and event signaling to decide the correct path automatically.

How do I recover stuck board tickets without losing local worktree changes?▼

Recovering stuck board tickets without losing changes requires applying concurrency guards and status checks to the active worktree. The process resumes work directly if changes are detected, otherwise it restarts.

Can I use board-api to update ticket status during a standalone local recovery?▼

Board-api updates are optional during recovery and apply to both board-enabled workflows and standalone local recoveries. This allows you to sync ticket status after successfully resuming or restarting the worktree.

What limitations should I expect when triggering stuck-ticket recovery for concurrent pipelines?▼

Stuck-ticket recovery for concurrent pipelines requires strict concurrency guards to prevent conflicts. If multiple processes attempt to resume or restart the same worktree simultaneously, the event signaling and cleanup steps may block execution.

Does the recover Skill require specific dependencies to restart empty worktrees?▼

Restarting empty worktrees requires no specific external dependencies, as the recovery process handles cleanup and event signaling internally. It automatically applies the restart logic when no usable worktree exists.