Agent-Skills.md

Agent Skills: Ruby on Rails

Conventions and best practices for building web applications with Ruby on Rails. Use when scaffolding Rails apps or generators, designing ActiveRecord models and migrations, wiring up Hotwire/Turbo/Stimulus interactivity, setting up background jobs or caching, or writing RSpec/Minitest coverage for Rails code.

UncategorizedID: mindrally/skills/ruby-rails

Repository

mindrally/skills
MindrallyLicense: Apache-2.0
25941

Install this agent skill to your local

pnpm dlx add-skill https://github.com/Mindrally/skills/tree/HEAD/ruby-rails

Skill Files

Browse the full folder contents for ruby-rails.

Download Skill

Loading file tree…

ruby-rails/SKILL.md

Skill Metadata

Name
ruby-rails
Description
"Conventions and best practices for building web applications with Ruby on Rails. Use when scaffolding Rails apps or generators, designing ActiveRecord models and migrations, wiring up Hotwire/Turbo/Stimulus interactivity, setting up background jobs or caching, or writing RSpec/Minitest coverage for Rails code."

Ruby on Rails

This skill covers Ruby and Ruby on Rails development, including Rails conventions, ActiveRecord patterns, frontend integration, security, and testing.

Core Principles

  • Write concise, idiomatic Ruby code with accurate examples
  • Adhere to Rails conventions (Convention over Configuration)
  • Follow the Ruby Style Guide for formatting consistency
  • Leverage Ruby 3.x features like pattern matching and endless methods
  • Prefer Rails command-line generators (bin/rails generate ...) over hand-written boilerplate — they keep files consistent with framework conventions and wire up tests/routes automatically
  • Use bin/dev to run the local development server (it boots the web process plus any configured watchers) and check logs after significant changes

Naming Conventions

  • Use snake_case for files, methods, and variables
  • Use CamelCase for classes and modules
  • Follow Rails naming conventions for models, controllers, views

Architecture & Performance

  • Utilize ActiveRecord for database operations with proper indexing
  • Use PostgreSQL in production, with proper indexes, connection pooling, and safe/reversible migrations
  • Implement eager loading to prevent N+1 query problems
  • Apply fragment caching and Russian Doll caching strategies
  • Use service objects for complex business logic
  • Follow MVC architecture strictly
  • Keep controllers RESTful and focused

Rails 8 Defaults

Rails 8 ships with a "no PaaS needed" stack — favor these built-ins over adding external services unless there's a specific reason not to:

  • Solid Queue — database-backed ActiveJob adapter; use it as the default background job backend instead of standing up Redis/Sidekiq unless job volume or feature needs (e.g., Sidekiq Pro batches) require it
  • Solid Cache — database-backed cache store; a reasonable default for Rails.cache without a separate Redis/Memcached deployment
  • Solid Cable — database-backed Action Cable adapter for WebSockets without Redis
  • Propshaft — the default asset pipeline; simpler than Sprockets, just serves fingerprinted files without a bundling step
  • Kamal — the default deployment tool for shipping containerized Rails apps to your own servers

Frontend & UI

  • Employ Hotwire (Turbo and Stimulus) for standard Rails interactivity and dynamic interactions without full page reloads
  • Reach for a Vite-managed JS pipeline (e.g., vite_rails) only when you actually need npm-managed JavaScript (React/Vue components, complex bundling) — don't add it by default
  • Design responsively with Tailwind CSS
  • Maintain DRY views through helpers and partials
  • Use ViewComponents for reusable UI components

Security

  • Implement authentication/authorization via Devise or Pundit
  • Use strong parameters in controllers to prevent mass assignment vulnerabilities
  • Sanitize user inputs appropriately
  • Use CSRF protection tokens
  • Implement proper session management

Testing

  • Write comprehensive Minitest or RSpec coverage for models, controllers, and integration flows, following TDD practices
  • Use FactoryBot for test data generation rather than fixtures
  • Mock external services; stub predefined return values
  • Use shared examples for common behaviors across different contexts
  • Ensure each test is independent; avoid shared state between tests

Best Practices

  • Keep controllers thin, models fat (but not too fat)
  • Use concerns for shared functionality
  • Implement background jobs with Sidekiq or ActiveJob
  • Use proper database migrations
  • Follow RESTful routing conventions