Rubigent

Rubigent is a lightweight Ruby library for building agentic AI systems. It provides a simple framework for creating and managing AI agents with multi-modal capabilities.

Features

• Create agents that can use LLM models
• Run agents with user prompts
• Capture and process agent responses
• Enable agents to utilize tools (if model permits)
• Support for structured outputs
• Track and report on agent execution with RunResponse and RunEvent

Installation

Add this line to your application's Gemfile:

gem 'rubigent'

And then execute:

$ bundle install

Or install it yourself as:

$ gem install rubigent

Usage

Basic Agent

require "rubigent"

# Create a simple agent
agent = Rubigent::Agent.new(
  model: Rubigent::Models::OpenAIChat.new(engine: "gpt-3.5-turbo"),
  description: "You are a helpful assistant."
)

# Run the agent with a prompt
response = agent.run("Hello, how are you?")
puts response.content

Advanced Run Response Tracking

require "rubigent"

# Create an agent
agent = Rubigent::Agent.new(
  model: Rubigent::Models::OpenAIChat.new(engine: "gpt-3.5-turbo"),
  description: "You are a helpful assistant."
)

# Create a start event
start_response = Rubigent::RunResponse.new(
  event: Rubigent::RunEvent::RUN_STARTED,
  content: "Starting agent run",
  run_id: "run-123",
  metrics: { start_time: Time.now.to_i }
)

# Run the agent
response = agent.run("Hello, how are you?")

# Add additional information to the response
response.event = Rubigent::RunEvent::RUN_COMPLETED
response.run_id = start_response.run_id
response.metrics = {
  start_time: start_response.metrics[:start_time],
  end_time: Time.now.to_i,
  duration: Time.now.to_i - start_response.metrics[:start_time]
}

# Add extra data
response.extra_data = Rubigent::RunResponseExtraData.new(
  reasoning_steps: [
    { step: 1, thought: "Analyzing the user prompt" },
    { step: 2, thought: "Generating a response" }
  ]
)

# Access response data
puts "Content: #{response.content}"
puts "Event: #{response.event}"
puts "Run ID: #{response.run_id}"
puts "Metrics: #{response.metrics.inspect}"
puts "Extra Data: #{response.extra_data.to_h.inspect}"

# Convert to JSON
puts response.to_json

Structured Output

require "rubigent"

# Define a response model
class MovieScript
  attr_accessor :setting, :ending, :genre, :name, :characters, :storyline

  def initialize(setting:, ending:, genre:, name:, characters:, storyline:)
    @setting = setting
    @ending = ending
    @genre = genre
    @name = name
    @characters = characters
    @storyline = storyline
  end
end

# Create an agent with structured output
agent = Rubigent::Agent.new(
  model: Rubigent::Models::OpenAIChat.new(engine: "gpt-4"),
  description: "You write movie scripts. Provide output as JSON.",
  response_model: MovieScript,
  structured_outputs: true
)

# Run the agent
response = agent.run("New York")
movie_script = response.content

puts "Setting: #{movie_script.setting}"
puts "Characters: #{movie_script.characters.join(', ')}"

Using Tools

require "rubigent"

# Create a tool
duckduckgo_tool = Rubigent::Tools::DuckDuckGoTool.new

# Create an agent with tools
agent = Rubigent::Agent.new(
  model: Rubigent::Models::OpenAIChat.new(engine: "gpt-4"),
  description: "You can use tools to find information.",
  tools: [duckduckgo_tool]
)

# Run the agent
response = agent.run("What's the capital of France?")
puts response.content

Cookbook

The Rubigent gem includes a cookbook with example recipes for using the framework. To explore the cookbook:

# List available examples
ruby cookbook/scripts/cookbook_runner.rb --list

# Run a specific example
ruby cookbook/scripts/cookbook_runner.rb --example movie_recommendation

Available examples include:

• async_basic - Demonstrates asynchronous agent usage
• structured_output - Shows how to use structured outputs
• tool_use - Illustrates tool integration
• movie_recommendation - A complete movie recommendation agent
• run_response_example - Demonstrates tracking and reporting on agent execution

Development

After checking out the repo, run bin/setup to install dependencies. Then, run rake test to run the tests. You can also run bin/console for an interactive prompt that will allow you to experiment.

To install this gem onto your local machine, run bundle exec rake install. To release a new version, update the version number in version.rb, and then run bundle exec rake release, which will create a git tag for the version, push git commits and the created tag, and push the .gem file to rubygems.org.

Contributing

Bug reports and pull requests are welcome on GitHub at https://github.com/[USERNAME]/rubigent. This project is intended to be a safe, welcoming space for collaboration, and contributors are expected to adhere to the code of conduct.

License

The gem is available as open source under the terms of the MIT License.

Code of Conduct

Everyone interacting in the Rubigent project's codebases, issue trackers, chat rooms and mailing lists is expected to follow the code of conduct.