I built a Rails engine for building and managing LLM-powered agents. It wraps RubyLLM and adds the production infrastructure I kept rebuilding across projects:

- Execution tracking with cost/token analytics - Reliability (retries, model fallbacks, circuit breakers) - Budget controls (daily/monthly limits per agent or tenant) - Workflow orchestration (pipelines, parallel, routers) - Real-time monitoring dashboard

Example:

    class SearchAgent < ApplicationAgent
        model "gpt-4o"
    
        reliability do
            retries max: 3, backoff: :exponential
            fallback_models "gpt-4o-mini", "claude-3-5-sonnet"
        end
    
        param :query, required: true
    
        def user_prompt
            "Extract search intent from: #{query}"
        end
    end

    result = SearchAgent.call(query: "red shoes under $50")
    result.content      # structured response
    result.total_cost   # $0.00025

GitHub: https://github.com/adham90/ruby_llm-agents

I've been using it in production for a few months. Happy to answer questions.