resuming

Resume interrupted HOTL workflows by loading sidecar state and re-verifying the last completed step.

25|2|Updated Mar 5, 2026
One-click install
npx skills add https://github.com/yimwoo/hotl-plugin --skill resuming-yimwoo
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: resuming
Source: https://github.com/yimwoo/hotl-plugin/tree/main/skills/resuming
Command: npx skills add https://github.com/yimwoo/hotl-plugin --skill resuming-yimwoo

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Resume an interrupted HOTL workflow by loading the sidecar state and re-validating the last step to continue execution.

Core Features & Use Cases

  • Sidecar state persistence and authoritative run metadata updates to recover from crashes.
  • Verify-first resume: re-checks completion status and continues from the next pending step.
  • Drift repair: aligns workflow checkboxes with sidecar state to maintain consistency after failures.
  • Robust resume flow: supports multiple resolution paths, stale run detection and safe user prompts.

Quick Start

Start HOTL again from a crash by specifying the run_id or workflow file.

Frequently Asked Questions about resuming

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

FAQPage Schema
How do I resume an interrupted workflow after a crash?

Resume an interrupted workflow by loading the sidecar state from the `.hotl/state` file and re-verifying the last completed step. This verify-first strategy ensures the workflow safely continues execution from the exact next pending step.

What is verify-first recovery for paused runs?

Verify-first recovery is a strategy that re-checks the completion status of the last step before continuing a paused run. It reads authoritative run metadata to safely resume execution and repairs any checkbox drift to maintain workflow consistency.

How do I repair workflow checkbox drift after a failure?

Repair checkbox drift by aligning the workflow checkboxes with the persisted sidecar state. The recovery process reads from `.hotl/state` to detect inconsistencies and automatically synchronizes the workflow state to ensure safe resumption.

Can I recover a crashed run across different sessions?

Yes, you can recover a crashed run across different sessions or restarts. The Skill loads the persisted sidecar state and preserves run metadata, enabling continuation from the last checkpoint regardless of when the interruption occurred.

How do I start HOTL again from a crash?

Start HOTL again from a crash by specifying the original run ID or the workflow file. The system loads the associated sidecar state, applies the verify-first strategy, and safely resumes the run from the next pending step.