dspy-ruby

name	dspy-ruby
user-invocable	true
description	This skill should be used when working with DSPy.rb, a Ruby framework for building type-safe, composable LLM applications. Use this when implementing predictable AI features, creating LLM signatures and modules, configuring language model providers (OpenAI, Anthropic, Gemini, Ollama), building agent systems with tools, optimizing prompts, or testing LLM-powered functionality in Ruby applications.
allowed-tools	["Bash","Read","Write","Edit","Grep","Glob"]
model	inherit

DSPy.rb Expert

Overview

DSPy.rb is a Ruby framework that enables developers to program LLMs, not prompt them. Instead of manually crafting prompts, define application requirements through type-safe, composable modules that can be tested, optimized, and version-controlled like regular code.

Mode Detection

Determine what the user needs, then read ONLY the relevant reference files before proceeding.

User Intent	Reference Files to Read
Signatures, modules, predictors, type safety	core-concepts.md
LLM provider setup, API keys, configuration	providers.md
Testing, optimization, observability	optimization.md
Getting started, Rails integration	quickstart.md
Common patterns, agents, pipelines, vision	patterns.md

All reference files are in ~/.claude/skills/dspy-ruby/references/.

---

Core Capabilities (Quick Reference)

1. Type-Safe Signatures

class EmailClassificationSignature < DSPy::Signature
  description "Classify customer support emails"

  input do
    const :email_subject, String
    const :email_body, String
  end

  output do
    const :category, T.enum(["Technical", "Billing", "General"])
    const :priority, T.enum(["Low", "Medium", "High"])
  end
end

Best practices: Clear descriptions, enums for constrained outputs, desc: parameter for fields, specific types over generic String. Full docs: core-concepts.md.

2. Composable Modules

class EmailProcessor < DSPy::Module
  def initialize
    super
    @classifier = DSPy::Predict.new(EmailClassificationSignature)
  end

  def forward(email_subject:, email_body:)
    @classifier.forward(email_subject: email_subject, email_body: email_body)
  end
end

Chain modules for complex workflows. Full docs: core-concepts.md.

3. Predictor Types

Predictor	Use For	Key Feature
Predict	Simple tasks, classification, extraction	Basic type-safe inference
ChainOfThought	Complex reasoning, analysis	Automatic reasoning step
ReAct	Tasks requiring external tools	Iterative tool-using agent
CodeAct	Tasks best solved with code	Dynamic code generation (dspy-code_act gem)

Full docs: core-concepts.md.

4. LLM Provider Configuration

# OpenAI
DSPy.configure { |c| c.lm = DSPy::LM.new('openai/gpt-4o-mini', api_key: ENV['OPENAI_API_KEY']) }

# Anthropic Claude
DSPy.configure { |c| c.lm = DSPy::LM.new('anthropic/claude-3-5-sonnet-20241022', api_key: ENV['ANTHROPIC_API_KEY']) }

# Google Gemini
DSPy.configure { |c| c.lm = DSPy::LM.new('gemini/gemini-1.5-pro', api_key: ENV['GOOGLE_API_KEY']) }

# Local Ollama (free, private)
DSPy.configure { |c| c.lm = DSPy::LM.new('ollama/llama3.1') }

Provider compatibility matrix:

Feature	OpenAI	Anthropic	Gemini	Ollama
Structured Output	yes	yes	yes	yes
Vision (Images)	yes	yes	yes	limited
Image URLs	yes	no	no	no
Tool Calling	yes	yes	yes	varies

Cost optimization: Dev = Ollama/gpt-4o-mini, Test = gpt-4o-mini (temp=0), Prod simple = gpt-4o-mini/haiku/flash, Prod complex = gpt-4o/sonnet/pro. Full docs: providers.md.

5. Testing

RSpec.describe EmailClassifier do
  before do
    DSPy.configure { |c| c.lm = DSPy::LM.new('openai/gpt-4o-mini', api_key: ENV['OPENAI_API_KEY']) }
  end

  it 'classifies technical emails correctly' do
    result = EmailClassifier.new.forward(email_subject: "Can't log in", email_body: "Unable to access account")
    expect(result[:category]).to eq('Technical')
  end
end

Full docs: optimization.md.

---

Reference Files Index

File	Content
core-concepts.md	Signatures, modules, predictors, multimodal support, best practices
providers.md	All LLM provider configs, compatibility matrix, cost optimization, troubleshooting
optimization.md	Testing patterns, MIPROv2 optimization, observability, monitoring
quickstart.md	Step-by-step new project setup, Rails integration
patterns.md	Multi-step pipelines, agents with tools, conditional routing, retry/fallback, vision, observability

Assets (templates for quick starts)

• assets/signature-template.rb -- Examples of signatures including basic, vision, sentiment, code gen
• assets/module-template.rb -- Module patterns including pipelines, agents, error handling, caching
• assets/config-template.rb -- Configuration for all providers, environments, observability, production

When to Use This Skill

• Implementing LLM-powered features in Ruby applications
• Creating type-safe interfaces for AI operations
• Building agent systems with tool usage
• Setting up or troubleshooting LLM providers
• Optimizing prompts and improving accuracy
• Testing LLM functionality
• Adding observability to AI applications
• Converting from manual prompt engineering to programmatic approach
• Debugging DSPy.rb code or configuration issues