code-prd

Transform PRD specifications into code with automated tests and reviews.

Updated Oct 25, 2025
One-click install
npx skills add https://github.com/navidemad/claude-code-plugins --skill code-prd
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: code-prd
Source: https://github.com/navidemad/claude-code-plugins/tree/main/prd-plugin/skills/code-prd
Command: npx skills add https://github.com/navidemad/claude-code-plugins --skill code-prd

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires jq, gh, and includes scripts (resource) and references (resource) components.

What problem does it solve?

Developers spend significant time manually implementing PRDs, writing tests, and performing code reviews. This Skill automates the structured development workflow, ensuring consistency, quality, and faster feature delivery. It acts as an AI pair programmer with process knowledge, reducing manual effort and cognitive load.

Core Features & Use Cases

  • Structured PRD Implementation: Guides Claude through substory-by-substory development, ensuring incremental progress and adherence to PRD requirements.
  • Automated Testing & Code Review: Automatically generates and runs tests, performs comprehensive code reviews, and even auto-fixes critical issues, ensuring high-quality code.
  • Context-Aware Development: Loads existing PRD context and project conventions (from CLAUDE.md) to maintain consistency, especially for feature expansions.
  • Use Case: A developer needs to implement a new feature defined in a PRD. They activate code-prd, and the AI guides them through each substory, writes the code, generates tests, reviews the changes, and asks for approval at each phase, significantly accelerating the development cycle.

Quick Start

Use the code-prd skill to implement the 'user-profile-core' PRD.

Frequently Asked Questions about code-prd

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

FAQPage Schema
How do I automate implementing features from a PRD with testing and code review?

Implementing PRD-driven features with automated testing and code review uses structured substory-by-substory development to transform specifications into production code. The code-prd Skill guides you through each phase, generates and runs tests automatically, performs comprehensive code review, and tracks progress with status markers (⏳ → 🔄 → ✅), reducing manual effort and ensuring consistency.

Can I use PRD-driven development with GitHub and automated tests?

PRD-driven development integrates GitHub workflows and automated testing through dependencies on gh and jq. The Skill generates tests, performs code reviews, and manages commits with concise messaging, making it compatible with GitHub-based development workflows and CI/CD pipelines.

What's the workflow for implementing multiple substories from a single PRD?

Substory-by-substory implementation loads context from your PRD and project conventions (CLAUDE.md), then guides incremental progress through each feature component. After each substory, automated tests run and code review occurs before approval, enabling iterative refinement with rollback checkpoints if needed.

Do I need to write tests manually when implementing features from a PRD?

Test generation is automated; the Skill generates and executes tests for each substory automatically. This eliminates manual test writing while ensuring code quality and catching issues early in the development cycle.

How does context loading from existing PRDs maintain consistency across feature expansions?

Context-aware development loads existing PRD specifications and project conventions to maintain consistency when expanding features. This ensures new code adheres to established patterns and requirements without manual context gathering.

What happens if automated code review finds issues during PRD implementation?

The Skill performs comprehensive code review after each phase and can auto-fix critical issues before requesting approval. Rollback checkpoints preserve your ability to revert changes if needed, enabling safe iteration.