Rails-AI
Star on GitHub
Modern Rails Development, Reimagined

Rails Domain Expertise for Claude Code

Opinionated Rails development with Workflow commands, domain skills, and battle-tested Superpowers workflows.

Explore on GitHub Read the Docs
100% Free
MIT License
Domain Skills
Rails Expertise
Team Rules
Enforced
3×
3× Faster
Development Speed

The Vision

Build Rails apps so autonomously they leave today's Next.js and React generators in the dust.

We're not playing catch-up—we're setting the pace and showing what LLM-driven development really looks like when Rails is in the driver's seat.

The Problem

Right now, LLMs excel at Next.js and Node code. Ask them to build a Rails app and you get... okay results. Ask for a Next.js app and you get production-ready code.

Why? Training data. JavaScript/TypeScript dominates the training sets. Rails, despite being more mature and productive, is underrepresented.

Why Rails is Perfect for AI

Rails is actually better suited for AI code generation:

  • ▸ Convention over Configuration — Less decision-making, more consistent patterns
  • ▸ Strong opinions — Clear right/wrong answers
  • ▸ Battle-tested — 20+ years of best practices
  • ▸ Monolithic by default — Simpler mental model
  • ▸ Solid Stack — Rails 8's batteries-included approach

Rails was built to make developers productive by removing decisions. That same philosophy makes it perfect for LLMs—fewer choices, more consistency, clearer patterns.

Two Great Frameworks

Rails-AI brings together battle-tested workflows and Rails expertise in a clean, layered architecture.

Superpowers

Process Layer

Universal workflows that work across any language or framework. The HOW of software development.

Brainstorming
Writing plans
Test-driven development
Systematic debugging
Code review
Verification gates

Rails-AI

Domain Layer

Rails-specific expertise and conventions. The WHAT you're building with Rails.

Domain skills
Team conventions
Workflow commands
Solid Stack
Minitest TDD
REST-only routing

Quick Start

Get started in under 2 minutes

Ready to install
Installation Commands 
# Install Superpowers (required dependency)
claude
/plugin marketplace add obra/superpowers
/plugin install superpowers

# Install Frontend Design (for ui skill)
/plugin marketplace add anthropics/claude-code
/plugin install frontend-design

# Install Rails-AI
/plugin marketplace add zerobearing2/rails-ai
/plugin install rails-ai

# Use workflow commands to build features
/rails-ai:feature add user authentication

Domain Skills

Each skill provides Rails-specific patterns, examples, and anti-patterns to avoid. Organized by domain for clear, focused expertise.

setup

Project validation, environment config, credentials, Docker, RuboCop. Coordinates with other skills for verification.

controllers

RESTful actions, strong parameters, concerns. Skinny controllers following Rails conventions.

debugging

Rails debugging tools + browser automation for JS console, visual issues, and HTML interactions.

hotwire

Turbo Drive, Frames, Streams, Morph. Stimulus controllers for interactive frontend.

jobs

SolidQueue, SolidCache, SolidCable. NO Redis/Sidekiq — Solid Stack only.

mailers

ActionMailer with async delivery via SolidQueue. Templates, previews, and testing patterns.

models

ActiveRecord patterns, validations, callbacks, scopes, associations, concerns, query objects.

security

CRITICAL — XSS, SQL injection, CSRF, file uploads, command injection prevention.

styling

Tailwind CSS utility-first framework and DaisyUI component library with theming support.

testing

TDD with Minitest, fixtures, model testing, mocking. NO RSpec — Minitest only.

ui

Powered by Frontend Design plugin. Designer→developer workflow for distinctive, production-grade interfaces.

performance

N+1 prevention, caching strategies, query optimization, Bullet integration.

Roadmap

✅ Complete

Phase 1 — Foundation

Initial skills, team rules, and testing framework. Established the foundation for Rails domain expertise in Claude Code.

🚧 Current

Phase 2 — Domain Workflows

Workflow commands replacing single architect coordinator, domain skills, deterministic Superpowers integration, and coordinator-only pattern with subagent dispatch for clean context windows.

🔬 Next

Phase 3 — Memory & Context

Index thousands of production Rails patterns, wire them into local memory + knowledge graph, and retrieve the right snippet at the right time with RAG + SQLite vector search.

🚀 Goal

Phase 4 — Fully Autonomous Rails

Ship end-to-end features—auth, payments, background jobs, admin, APIs—complete with tests and security. Target: 85%+ first-pass success rate.

Frequently Asked Questions

Rails-AI is a Claude Code plugin that adds Rails domain expertise through a layered architecture. Workflow commands (/rails-ai:setup, plan, feature, refactor, debug, review) load the appropriate Superpowers workflows (process layer) and Rails-AI skills (domain layer) for each task.

Yes! Rails-AI is completely free and open source under the MIT license. You'll need a Claude Code subscription to use the plugin, but Rails-AI itself is free forever.

Rails-AI is optimized for Rails 8+ and includes built-in knowledge of the Solid Stack (SolidQueue, SolidCache, SolidCable), Hotwire (Turbo + Stimulus), Tailwind CSS, and modern Rails patterns following 37signals philosophy.

Installation is simple:

  1. Install Superpowers plugin: 
    /plugin marketplace add obra/superpowers
/plugin install superpowers
  2. Install Frontend Design plugin: 
    /plugin marketplace add anthropics/claude-code
/plugin install frontend-design
  3. Install Rails-AI: 
    /plugin marketplace add zerobearing2/rails-ai
/plugin install rails-ai
  4. Use workflow commands like /rails-ai:feature to start building.

Rails-AI includes focused domain skills:

  • setup - Environment config, credentials, Docker, RuboCop
  • models - ActiveRecord patterns, validations, associations, callbacks
  • controllers - RESTful actions, strong parameters, concerns
  • ui - Frontend Design plugin integration for distinctive interfaces
  • hotwire - Turbo (Drive, Frames, Streams, Morph) + Stimulus
  • styling - Tailwind CSS + DaisyUI theming
  • testing - TDD with Minitest, fixtures, mocking
  • security - XSS, SQL injection, CSRF, file uploads
  • jobs - SolidQueue, SolidCache, SolidCable (NO Redis/Sidekiq)
  • mailers - ActionMailer, async delivery, templates
  • debugging - Rails tools + browser debugging
  • performance - N+1 prevention, caching, query optimization

Rails-AI is built on Superpowers, a plugin that provides universal workflows (brainstorming, planning, TDD, debugging, code review). Rails-AI adds the Rails-specific domain knowledge on top. Together: Superpowers = HOW to work, Rails-AI = WHAT you're building.

Rails-AI enforces team conventions with critical rules that are non-negotiable:

  • NEVER Sidekiq/Redis SolidQueue/SolidCache
  • NEVER RSpec Minitest only
  • NEVER custom routes RESTful resources
  • NEVER skip TDD RED-GREEN-REFACTOR
  • NEVER merge without review Draft PRs
  • NEVER WebMock bypass Mock all HTTP

These rules enforce 37signals philosophy and Rails conventions automatically.

Yes! Rails-AI works with both new and existing Rails applications. The architect analyzes your existing codebase structure and follows your established patterns. It understands Rails architecture and will integrate seamlessly with your current models, controllers, views, and tests.

Join the Rails Renaissance

This is bigger than a plugin. We're proving Rails can lead in the AI era.

Rails Developers

Try the architect. Break it. Tell us what's missing. Your real-world usage drives what we build.

AI/ML Engineers

We need help with Phase 3 (RAG) and Phase 4 (autonomous delivery). Indexing, evaluation, retrieval quality.

Rails Advocates

Star the repo. Spread the word. Rails deserves world-class AI tooling. Help us build it.

Contribute on GitHub

The goal isn't just good—it's to make Rails the obvious choice for AI-assisted development.