deep-execute

Processes bulk development task queues concurrently with Git conflict resolution and recovery support.

Updated Jan 23, 2026
One-click install
npx skills add https://github.com/marcusgoll/deep-loop-plugin --skill deep-execute
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: deep-execute
Source: https://github.com/marcusgoll/deep-loop-plugin/tree/main/skills/deep-execute
Command: npx skills add https://github.com/marcusgoll/deep-loop-plugin --skill deep-execute

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill eliminates the slow, error-prone process of manually processing queued development tasks one at a time, drastically reducing execution time for bulk work batches.

Core Features & Use Cases

  • Concurrent Worker Orchestration: Launch up to 8 parallel Claude CLI workers to process tasks from your .deep/tasks.md queue simultaneously, with configurable worker counts.
  • Claim-Based Coordination: Built-in lock system prevents duplicate work across parallel workers, with automatic stale claim recovery for interrupted runs.
  • Git Conflict Handling: Auto-resolves push conflicts, creates dedicated recovery branches for blocked tasks, and retries failed pushes up to 3 times before marking tasks as blocked.
  • Use Case: Ideal for development teams running bulk TDD task batches, overnight autonomous work sessions, or recovering from interrupted task runs without manual rework.

Quick Start

Use the deep-execute skill to process all pending tasks in your .deep/tasks.md queue with 3 parallel workers.

Frequently Asked Questions about deep-execute

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

FAQPage Schema
How do I run parallel development tasks to speed up bulk TDD batches?

Parallel development tasks accelerate bulk TDD batches by launching up to 8 concurrent Claude CLI workers to process your task queue simultaneously. This eliminates slow sequential execution and drastically reduces overall batch processing time.

How does claim-based coordination prevent duplicate work during concurrent task execution?

Claim-based coordination uses a built-in lock system to prevent duplicate work across parallel workers. It automatically manages task claiming from the queue and includes stale claim recovery to resume processing if an interrupted run leaves locks behind.

What's the best way to handle git conflicts when running autonomous development tasks overnight?

Handling git conflicts during autonomous overnight runs requires automatic push conflict resolution. The system auto-resolves conflicts, creates dedicated recovery branches for blocked tasks, and retries failed pushes up to 3 times before marking tasks as blocked.

Can I recover crashed bulk task processing runs without manual rework?

Crash recovery for interrupted bulk task processing runs is supported through automatic stale claim recovery. This mechanism detects interrupted workers, clears stale locks, and re-queues tasks so processing resumes without manual rework.

How many parallel workers can I configure for concurrent task orchestration?

Concurrent task orchestration supports configurable worker counts up to a maximum of 8 parallel Claude CLI workers. This allows you to balance processing speed against system resource limits when executing your task queue.