ultrapilot

Partition code ownership across up to five workers for parallel autopilot execution.

2|Updated Feb 3, 2026
One-click install
npx skills add https://github.com/chanhee-kang/oh-my-gemini --skill ultrapilot
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ultrapilot
Source: https://github.com/chanhee-kang/oh-my-gemini/tree/main/skills/ultrapilot
Command: npx skills add https://github.com/chanhee-kang/oh-my-gemini --skill ultrapilot

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Ultrapilot solves the challenge of accelerating large, multi-component tasks by partitioning work into independent workers with exclusive file ownership, enabling safe parallel execution.

Core Features & Use Cases

  • Parallel task decomposition: splits a complex task into multiple subtasks that can run concurrently without interfering file changes.
  • Conflict-aware integration: coordinates shared files and boundary imports to guarantee a clean merge during final integration.
  • Configurable orchestration: supports a cap of five workers, validation rounds, and flexible conflict handling policies.
  • Use Case: refactor a full-stack feature across frontend, backend, and database modules by assigning each module to a separate worker and then integrating results in a coordinated phase.

Quick Start

/oh-my-gemini:ultrapilot Build a multi-service feature with parallel modules

Frequently Asked Questions about ultrapilot

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

FAQPage Schema
How do I run parallel code refactoring across multiple modules without file conflicts?

Parallel code refactoring across multiple modules requires partitioning code ownership across workers. This skill enables concurrent execution by enforcing exclusive file ownership and coordinating shared files through a central integration phase to prevent conflicts.

What is the best way to orchestrate full-stack feature development across frontend, backend, and database layers?

Orchestrating full-stack feature development is best handled by assigning independent modules like frontend, backend, and database to separate concurrent workers. This skill caps execution at five workers and coordinates boundary imports during a final integration phase.

How does file ownership work during concurrent multi-component development?

File ownership during concurrent multi-component development works by granting exclusive write access to specific files for each worker. Shared files and boundary imports are coordinated through a central integration phase to guarantee a clean merge.

Can I configure validation rounds and conflict policies for parallel task decomposition?

Yes, you can configure validation rounds and flexible conflict handling policies for parallel task decomposition. The orchestration supports a cap of five concurrent workers and allows you to define how conflicts are managed during the integration phase.

How many concurrent workers can I use for multi-module orchestration?

You can use up to five concurrent workers for multi-module orchestration. This limit ensures safe parallel execution and manageable conflict resolution when integrating shared files and boundary imports across independent modules.

When should I avoid using parallel autopilot for task decomposition?

You should avoid parallel autopilot execution when your development task cannot be partitioned into independent modules with exclusive file ownership. Tasks requiring heavy modifications to shared files across multiple components will face bottlenecking during the integration phase.