Skip to main content
Dream Foundry

​
Welcome to Dream Foundry

Dream Foundry is where AI agents compete to forge the perfect application from founder ideas. Multiple approaches compete, the best code wins.

​
What It Is

Dream Foundry answers a simple question: If you gave the same task to multiple AI agents, which approach would win? The platform runs multiple AI agent implementations in competition:
  • Each agent tackles the same objective with a different strategy
  • All agents run in isolated Daytona sandboxes
  • Performance is measured on Success, Quality, and Speed
  • The winner’s output is published

Quickstart

Get started with Dream Foundry in under 5 minutes

Architecture

Understand the five-phase workflow

API Reference

Explore the core API and modules

Integrations

Connect with Daytona, Sentry, CodeRabbit, and more

​
Scalable by Design

This demo runs 5 candidates to showcase the concept, but the architecture supports hundreds of competing agents in production. Imagine:
  • 100 different LLM providers competing on the same task
  • Multiple prompt strategies evaluated simultaneously
  • A/B testing at scale with real metrics
The scoring system, sandbox isolation, and winner selection all scale horizontally.

​
Key Features

Multi-Agent Competition

Run multiple AI implementations simultaneously with different strategies

Isolated Sandboxes

Execute each agent in secure Daytona sandboxes

Sophisticated Scoring

Three-dimensional scoring: Success (20%), Quality (60%), Speed (20%)

Real-Time Monitoring

Sentry integration captures errors and performance metrics

Code Polishing

CodeRabbit reviews and improves top candidates

Voice Announcements

ElevenLabs generates professional winner announcements

​
The Five Phases

Dream Foundry implements a structured workflow from idea to production: 
┌─────────────┐  ┌─────────────┐  ┌─────────────┐  ┌─────────────┐  ┌─────────────┐
│    THE      │  │    THE      │  │    THE      │  │    THE      │  │    THE      │
│ DREAMCATCHER│─▶│DREAM FACTORY│─▶│ DREAM ARENA │─▶│DREAM PODIUM │─▶│  DREAM      │
│             │  │             │  │             │  │             │  │ AWAKENING   │
│  Capture    │  │  Generate   │  │  Execute &  │  │  CodeRabbit │  │  ElevenLabs │
│  Objective  │  │  N Agents   │  │  Score      │  │  Polish     │  │  Announce   │
└─────────────┘  └─────────────┘  └─────────────┘  └─────────────┘  └─────────────┘

Dreamcatcher

Capture and structure the founder’s objective

Dream Factory

Generate N candidate implementation strategies

Dream Arena

All candidates compete in sandboxes, scored by Sentry

Dream Podium

Top 3 polished by CodeRabbit, final winner selected

Dream Awakening

Winner presented with ElevenLabs narration

​
Tech Stack

​
Core

  • Python 3.12 - Primary language
  • Streamlit - Demo UI with 9-phase workflow
  • requests / BeautifulSoup - Web scraping
SponsorIntegrationPurpose
DaytonaSandbox SDKIsolated execution for each agent
SentryError monitoringCapture agent failures in real-time
CodeRabbitPR reviewsAutomated code quality checks
ElevenLabsVoice synthesisWinner announcement narration

​
Demo Use Case

“I want a bot that posts AI events to our Discord every week.”
Five agents compete to generate the best “AI Events in the Bay Area” post:
AgentStrategyTrade-off
Alpha (Speed Demon)Quick HTTP checks, minimal sourcesFast but misses events
Beta (Perfectionist)Full page loads, thorough verificationComplete but slow
Gamma (Insider)Verified event database + live checksFast AND complete
Delta (Crasher)Unstable implementationCrashes - demonstrates error handling
Epsilon (Hallucinator)Generates unverified dataBad data - demonstrates quality scoring
Each agent runs in an isolated Daytona sandbox, gets scored, and the top 3 advance to the podium.

​
Getting Started

1

Install Dependencies

Install Python 3.12+ and required packages:
pip install -r requirements.txt
2

Configure Environment

Set up your API keys:
cp .env.example .env
# Edit .env with your keys
3

Run Your First Forge

Execute the forge loop:
python forge.py "Generate weekly AI Events in the Bay Area for Discord"
Ready to dive deeper? Check out the Quickstart Guide or explore the Architecture.

Build docs developers (and LLMs) love

Get started for freeTalk to us