resume-mission

To resume an interrupted AI agent mission, the skill scans the `.mission/` directory for durable state files to reconstruct the mission context. It re-enters the canonical mission loop at the exact planning, execution, or validation phase where the workflow was previously paused.

State reconstruction for paused AI workflows works by reading durable state files stored in `.mission/<slug>/` directories. These files capture the mission context, allowing the state management process to rebuild the exact workflow stage and resume execution seamlessly.

When multiple active AI missions exist, the skill performs intelligent mission selection by prioritizing active missions over terminal ones. It identifies and selects the most recently active mission candidate to ensure you re-enter the correct workflow context.

Handling missing or corrupted state files during mission resumption requires robust error handling built into the state reconstruction process. The skill validates the `.mission/` directory contents to gracefully manage corrupted state files without crashing the AI workflow.

Yes, you can re-enter the execution phase directly without repeating the planning stage. The skill ensures resumption point accuracy by reconstructing the mission state and jumping directly into the canonical mission loop at the precise phase where it was interrupted.

No, you do not need to manually retrace steps to resume a paused AI coding task. The skill scans durable state files to automatically rebuild the mission context, preventing loss of progress and allowing you to pick up exactly where you left off.