dev

Orchestrate software delivery from requirements to reviewed, tested code.

88|6|Updated Mar 25, 2026
One-click install
npx skills add https://github.com/heliohq/ship --skill dev-heliohq
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: dev
Source: https://github.com/heliohq/ship/tree/main/skills/dev
Command: npx skills add https://github.com/heliohq/ship --skill dev-heliohq

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Ship coordinates disciplined AI software delivery so agents can implement, review, test, and hand off code without skipping quality gates or losing state.

Core Features & Use Cases

  • Routes work into single-story tasks, bundled phases, or full end-to-end delivery flows.
  • Keeps requirements, implementation state, and outputs on disk so each phase stays isolated and evidence-backed.
  • Supports common engineering workflows like building from a spec, reviewing changes, validating QA findings, refactoring safely, and recovering broken PRs.

Quick Start

Use the dev skill to implement the provided story with TDD, peer review, and a final regression test.

Frequently Asked Questions about dev

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

FAQPage Schema
How do I automate TDD and peer code review for AI-generated code?

You can automate TDD and peer code review by orchestrating AI delivery workflows that enforce test-driven development, peer cross-validation, and evidence-backed verification before handing off code. This ensures agents implement features without skipping quality gates.

What is the best way to manage AI software delivery state across isolated engineering phases?

The best way to manage AI software delivery state is keeping requirements, implementation state, and outputs on disk so each engineering phase stays isolated, evidence-backed, and recoverable across story implementation or bundled delivery flows.

Can I use isolated phase context to recover broken pull requests and validate QA findings?

Yes, you can use isolated phase context to recover broken PRs and validate QA findings by routing work into single-story tasks or bundled phases that maintain disk-based state for reliable regression testing and peer review.

Does this AI workflow approach support refactoring and phase-bundled delivery?

This AI workflow approach supports refactoring and phase-bundled delivery by routing raw requirements into implemented, reviewed, and tested code through isolated phase context that enforces TDD and peer cross-validation for reliable releases.

Why does my AI coding agent skip quality gates during story implementation?

AI coding agents skip quality gates when lacking orchestrated software delivery workflows that enforce TDD, evidence-backed verification, and peer cross-validation. Disk-based state and isolated phase context prevent losing progress and skipping reviews.