resume

Session resume for in_progress stories works by querying the hub API to detect active work, evaluating lock status, reading the execution plan and log to find the correct resume point, and updating the lock to proceed safely from the first incomplete task.

Active lock conflicts are handled by enforcing lock handoffs and safe resume decisions. The system detects when a different agent holds an active lock, reports the conflict status, and guides the user to pause rather than overwriting concurrent execution.

Stale lock auto-takeover triggers when a lock is older than four hours. The system evaluates lock age during the resume state check, automatically takes over stale locks, and updates the lock so the current agent can proceed with execution.

Missing execution data triggers failure handling guidance during the resume state report. The system detects missing execution plans or logs from the hub API and provides specific error handling instructions rather than attempting a blind resume.

A hub API connection is required to manage story locks and execution logs. The system queries the hub API to read the execution plan and log, evaluate lock status, and update locks when resuming in_progress stories.

The best way to handle conflicting lock scenarios is to evaluate lock status via the hub API before proceeding. The system distinguishes between no lock, stale locks older than four hours, and active locks by the same or different agents to guide safe resume decisions.