Rails Generator for RubyLLM Models

CLAUDE.md

@@ -0,0 +1,24 @@
+# RubyLLM Development Guidelines
+
+## Commands
+- Run all tests: `bundle exec rspec`
+- Run single test: `bundle exec rspec spec/path/to/file_spec.rb`
+- Run specific test: `bundle exec rspec spec/path/to/file_spec.rb:LINE_NUMBER`
+- Code style check: `bundle exec rubocop`
+- Auto-fix style issues: `bundle exec rubocop -A`
+- Record VCR cassettes: `bundle exec rake vcr:record[provider]` (where provider is openai, anthropic, gemini, etc.)
+
+## Style Guidelines
+- Follow Standard Ruby style (https://github.com/testdouble/standard)
+- Use frozen_string_literal comment at top of files
+- Add YARD documentation comments for public methods
+- Use proper error handling with specific error classes
+- Follow consistent model naming conventions across providers
+- Keep normalized model IDs (separate model from provider)
+- Use consistent parameter naming across providers
+
+## Testing Practices
+- Tests automatically create VCR cassettes based on their descriptions
+- Use specific, descriptive test descriptions
+- Check VCR cassettes for sensitive information
+- Write tests for all new features and bug fixes

CONTRIBUTING.md

@@ -133,6 +133,9 @@ bundle exec rspec
 
 # Run a specific test file
 bundle exec rspec spec/ruby_llm/chat_spec.rb
+
+# Run generator template tests
+bundle exec rspec spec/lib/generators/ruby_llm/template_files_spec.rb
 ```
 
 ### Recording VCR Cassettes

README.md

@@ -137,11 +137,37 @@ chat.with_model('gemini-2.0-flash').ask "What's your favorite algorithm?"
 
 ## Rails integration that makes sense
 
+### Option 1: Create new models with the generator
+
+Simply run the generator to set up all required models:
+
+```bash
+rails generate ruby_llm:install
+```
+
+This creates all necessary migrations, models, and database tables. Then just use them:
+
+```ruby
+# In your controller
+chat = Chat.create!(model_id: "gpt-4o-mini")
+chat.ask("What's your favorite Ruby gem?") do |chunk|
+  Turbo::StreamsChannel.broadcast_append_to(
+    chat, target: "response", partial: "messages/chunk", locals: { chunk: chunk }
+  )
+end
+
+# That's it - chat history is automatically saved
+```
+
+### Option 2: Add to your existing models
+
+Or, add RubyLLM to your existing ActiveRecord models:
+
 ```ruby
 # app/models/chat.rb
 class Chat < ApplicationRecord
   acts_as_chat
-
+  
   # Works great with Turbo
   broadcasts_to ->(chat) { "chat_#{chat.id}" }
 end

docs/guides/rails.md

@@ -12,9 +12,32 @@ RubyLLM provides seamless integration with Rails through ActiveRecord models. Th
 
 ## Setup
 
-### 1. Create Migrations
+### Using the Generator (Recommended)
 
-First, create the necessary tables in your database:
+The easiest way to set up RubyLLM with Rails is to use the built-in generator:
+
+```bash
+rails generate ruby_llm:install
+```
+
+This will automatically:
+1. Create the necessary migrations for chats, messages, and tool calls
+2. Create model files with appropriate `acts_as_*` methods
+3. Set up proper relationships between models
+
+After running the generator, simply run the migrations:
+
+```bash
+rails db:migrate
+```
+
+### Manual Setup
+
+If you prefer to set up manually or need to customize the implementation, follow these steps:
+
+#### 1. Create Migrations
+
+Create the necessary tables in your database:
 
 ```ruby
 # db/migrate/YYYYMMDDHHMMSS_create_chats.rb
@@ -65,7 +88,7 @@ Run the migrations:
 rails db:migrate
 ```
 
-### 2. Set Up Models
+#### 2. Set Up Models
 
 Create the model classes:
 

lib/generators/ruby_llm/install/templates/chat_model.rb

@@ -0,0 +1,5 @@
+# frozen_string_literal: true
+
+class Chat < ApplicationRecord
+  acts_as_chat
+end

lib/generators/ruby_llm/install/templates/create_chats_migration.rb

@@ -0,0 +1,10 @@
+# frozen_string_literal: true
+
+class CreateChats < ActiveRecord::Migration<%= migration_version %>
+  def change
+    create_table :chats do |t|
+      t.string :model_id
+      t.timestamps
+    end
+  end
+end

lib/generators/ruby_llm/install/templates/create_messages_migration.rb

@@ -0,0 +1,18 @@
+# frozen_string_literal: true
+
+# This migration must be run AFTER create_chats and create_tool_calls migrations
+# to ensure proper foreign key references
+class CreateMessages < ActiveRecord::Migration<%= migration_version %>
+  def change
+    create_table :messages do |t|
+      t.references :chat, null: false, foreign_key: true
+      t.string :role
+      t.text :content
+      t.string :model_id
+      t.integer :input_tokens
+      t.integer :output_tokens
+      t.references :tool_call, foreign_key: true
+      t.timestamps
+    end
+  end
+end

lib/generators/ruby_llm/install/templates/create_tool_calls_migration.rb

@@ -0,0 +1,24 @@
+# frozen_string_literal: true
+
+# Migration for creating tool_calls table with database-specific JSON handling
+class CreateToolCalls < ActiveRecord::Migration<%= migration_version %>
+  def change
+    create_table :tool_calls do |t|
+      # No reference to message to avoid circular references
+      # Messages will reference tool_calls, not the other way around
+      t.string :tool_call_id, null: false
+      t.string :name, null: false
+
+      # Use the appropriate JSON column type for the database
+      if postgresql?
+        t.jsonb :arguments, default: {}
+      else
+        t.json :arguments, default: {}
+      end
+
+      t.timestamps
+    end
+
+    add_index :tool_calls, :tool_call_id
+  end
+end

lib/generators/ruby_llm/install/templates/initializer.rb

@@ -0,0 +1,16 @@
+# frozen_string_literal: true
+
+# RubyLLM configuration
+RubyLLM.configure do |config|
+  # Set your API keys here or use environment variables
+  # config.openai_api_key = ENV["OPENAI_API_KEY"]
+  # config.anthropic_api_key = ENV["ANTHROPIC_API_KEY"]
+  # config.gemini_api_key = ENV["GEMINI_API_KEY"]
+  # config.deepseek_api_key = ENV["DEEPSEEK_API_KEY"]
+
+  # Uncomment to set a default model
+  # config.default_model = "gpt-4o-mini"
+
+  # Uncomment to set default options
+  # config.default_options = { temperature: 0.7 }
+end

lib/generators/ruby_llm/install/templates/message_model.rb

@@ -0,0 +1,5 @@
+# frozen_string_literal: true
+
+class Message < ApplicationRecord
+  acts_as_message
+end

lib/generators/ruby_llm/install/templates/tool_call_model.rb

@@ -0,0 +1,5 @@
+# frozen_string_literal: true
+
+class ToolCall < ApplicationRecord
+  acts_as_tool_call
+end

lib/generators/ruby_llm/install_generator.rb

@@ -0,0 +1,63 @@
+# frozen_string_literal: true
+
+require "rails/generators"
+require "rails/generators/active_record"
+
+module RubyLLM
+  # Generator for RubyLLM Rails models and migrations
+  class InstallGenerator < Rails::Generators::Base
+    include Rails::Generators::Migration
+
+    source_root File.expand_path("install/templates", __dir__)
+
+    desc "Creates model files for Chat, Message, and ToolCall, and creates migrations for RubyLLM Rails integration"
+
+    def self.next_migration_number(dirname)
+      ActiveRecord::Generators::Base.next_migration_number(dirname)
+    end
+
+    def migration_version
+      "[#{Rails::VERSION::MAJOR}.#{Rails::VERSION::MINOR}]"
+    end
+
+    def postgresql?
+      ActiveRecord::Base.connection.adapter_name.downcase.include?("postgresql")
+    rescue
+      false
+    end
+
+    def create_migration_files
+      # Create migrations in the correct order with sequential timestamps
+      # to ensure proper foreign key references:
+      # 1. First create chats (no dependencies)
+      # 2. Then create tool_calls (will be referenced by messages)
+      # 3. Finally create messages (depends on both chats and tool_calls)
+
+      # Use a fixed timestamp for testing and to ensure they're sequential
+      @migration_number = Time.now.utc.strftime("%Y%m%d%H%M%S")
+      migration_template "create_chats_migration.rb", "db/migrate/create_chats.rb"
+
+      # Increment timestamp for the next migration
+      @migration_number = (@migration_number.to_i + 1).to_s
+      migration_template "create_tool_calls_migration.rb", "db/migrate/create_tool_calls.rb"
+
+      # Increment timestamp again for the final migration
+      @migration_number = (@migration_number.to_i + 2).to_s
+      migration_template "create_messages_migration.rb", "db/migrate/create_messages.rb"
+    end
+
+    def create_model_files
+      template "chat_model.rb", "app/models/chat.rb"
+      template "message_model.rb", "app/models/message.rb"
+      template "tool_call_model.rb", "app/models/tool_call.rb"
+    end
+
+    def create_initializer
+      template "initializer.rb", "config/initializers/ruby_llm.rb"
+    end
+
+    def show_readme
+      readme "README.md"
+    end
+  end
+end

lib/ruby_llm.rb

@@ -20,6 +20,7 @@
 loader.ignore("#{__dir__}/tasks")
 loader.ignore("#{__dir__}/ruby_llm/railtie")
 loader.ignore("#{__dir__}/ruby_llm/active_record")
+loader.ignore("#{__dir__}/generators")
 loader.setup
 
 # A delightful Ruby interface to modern AI language models.

lib/ruby_llm/railtie.rb

@@ -8,5 +8,15 @@ class Railtie < Rails::Railtie
         include RubyLLM::ActiveRecord::ActsAs
       end
     end
+
+    # Include rake tasks if applicable
+    rake_tasks do
+      # Task definitions go here if needed
+    end
+
+    # Register generators
+    generators do
+      require 'generators/ruby_llm/install_generator'
+    end
   end
 end