loop

Automate large project execution with crash recovery and per-task isolation.

2|Updated Feb 4, 2026
One-click install
npx skills add https://github.com/lhaig/loop --skill loop
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: loop
Source: https://github.com/lhaig/loop/tree/main/loop
Command: npx skills add https://github.com/lhaig/loop --skill loop

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Loop automates large project execution with crash recovery and persistent state, enabling you to resume workflows after interruptions without losing progress.

Core Features & Use Cases

  • Autonomous orchestration of long-running features with per-task subagents.
  • Crash recovery and session persistence via .loop state files.
  • Interactive planning, PRD import, and continued loop execution across sessions.

Quick Start

Start by initializing a loop-enabled project: say "loop plan" to begin planning, or "loop import tasks/prd-<name>.md" to import a PRD, then "continue loop" to execute tasks.

Frequently Asked Questions about loop

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

FAQPage Schema
How do I automate large project execution across multiple sessions?

Automate large project execution across multiple sessions by orchestrating tasks with persistent state files, enabling you to resume workflows seamlessly after interruptions or crashes.

What is the best way to resume a workflow after a crash without losing progress?

Resume a workflow after a crash by restoring execution state from local checkpoint files and progress logs, ensuring no completed task is lost or repeated during the session.

How do I import a PRD and break it down into isolated tasks for autonomous execution?

Import a PRD by passing the markdown file path to the planning command, which parses the document to generate an interactive task list for autonomous subagent execution.

Can I use crash recovery and task isolation for workflows involving 10+ tasks?

Yes, crash recovery and per-task isolation are explicitly designed for workflows involving 10+ tasks, ensuring each task executes independently while maintaining safe checkpoint logs.

How does state persistence work when managing long-running project automation?

State persistence works by continuously saving execution progress in a local folder, enforcing safe checkpoints via automated commits and progress logs to prevent data loss.

Do I need a PreCompact hook to reset context during autonomous task planning?

A PreCompact hook is optional for automatic context resets during task planning, but the core workflow functions autonomously using state persistence and crash recovery without requiring it.