autopilot

Orchestrate a strict autonomous loop from planning through implementation and code review.

Updated Apr 20, 2026
One-click install
npx skills add https://github.com/hotaq/Sprite_harmess --skill autopilot-hotaq
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: autopilot
Source: https://github.com/hotaq/Sprite_harmess/tree/main/.codex/skills/autopilot
Command: npx skills add https://github.com/hotaq/Sprite_harmess --skill autopilot-hotaq

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Autopilot removes the friction of coordinating complex software work by taking a task from planning through implementation and review in a strict, repeatable loop.

Core Features & Use Cases

  • Strict phased workflow: Follows the exact ralplan -> ralph -> code-review sequence without inventing extra lifecycle stages.
  • Review-driven iteration: If review is not clean, it returns to planning with findings and continues until the work is ready or blocked.
  • Stateful task execution: Persists context, handoff artifacts, and phase status so long-running tasks can resume safely.
  • Use case: A developer hands over an issue, PRD, or feature request and lets the skill plan the work, implement the changes, and keep iterating until review approval.

Quick Start

Ask Autopilot to take your task from planning through implementation and code review using the strict autonomous delivery loop.

Frequently Asked Questions about autopilot

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

FAQPage Schema
How do I automate software delivery from task planning through code review?

Autonomous software delivery is orchestrated by running a strict plan-to-review loop that handles planning, implementation, and code review in sequence. The workflow persists task context and phase status so long-running development tasks can resume safely without losing progress.

What happens if an autonomous code review fails during the delivery loop?

Failed code review triggers review-driven iteration by returning the task to the planning phase with the review findings. The autonomous loop continues to plan, implement, and verify the changes until the code review is clean or the task becomes blocked.

How do I run a stateful workflow for long-running implementation tasks?

Stateful workflow execution persists context, handoff artifacts, and phase status across the planning, implementation, and review phases. This allows long-running autonomous delivery tasks to resume safely without losing their execution state.

Do I need approved planning artifacts to start autonomous implementation?

Approved planning artifacts are required to start the autonomous implementation workflow. The delivery loop requires approved ralplan artifacts, implementation evidence, and a clean code-review verdict to complete the phased execution successfully.

What is the best way to run hands-off software development with review-driven iteration?

Hands-off software delivery is achieved by initiating a strict autonomous loop that takes an issue or feature request from planning through implementation. It automatically iterates based on code review findings until the work is ready for approval.

When should I not use an autonomous delivery loop for software engineering?

An autonomous delivery loop should not be used when a task requires manual intervention outside the strict plan-to-review sequence. The workflow is designed for hands-off execution and cannot invent extra lifecycle stages beyond planning, implementation, and review.