What to Build

Build an AI-Powered Application Using Prompt-Driven Development

Create a project that demonstrates the power of Prompt-Driven Development (PDD) - where prompts are the source of truth and code is a regenerable artifact.

Eligible Project Types

Agentic Applications

• AI agents that take real-world actions via Toolhouse

• Automated workflows that interact with external APIs and services

• Multi-step task automation with LLM-powered decision making

Developer Tools

• Tools that enhance the PDD workflow

• Code generation utilities that follow PDD principles

• Testing frameworks that support test accumulation 

Enterprise Solutions

• Business process automation using prompt-defined logic

• Customer-facing AI applications with maintainable architecture

• Data processing pipelines defined by regenerable prompts 

Creative Applications

• Interactive applications powered by AI agents

• Content generation tools with structured prompt workflows

• Educational platforms leveraging LLM capabilities

Required Technologies

Prompt-Driven Development (PDD) - Your project must demonstrate PDD principles:

• Prompts define behavior; code is generated

• Changes made by modifying prompts and regenerating (not patching)

• Intent preserved in prompts, not buried in code

Toolhouse.ai (Recommended) - Give your LLMs the ability to:

• Use tools and interact with the internet

• Execute real-world actions

• Access external services and APIs

Rtrvr.ai (Recommended) - Visit Resources for more information

ElevenLabs (Recommended) - Visit Resources for more information

Why PDD?

From the PDD whitepaper benchmarks:

• 55% reduction in active developer time vs. interactive patching

• 100% success rate in complex tasks vs. 46.67% for traditional approaches

• 10x lower cost per successful operation ($0.08 vs $0.77)

PDD treats prompts like molds in manufacturing - invest in the specification once, regenerate the artifact as needed. Your tests become "mold walls" that permanently prevent bugs from recurring.

What to Submit

Required Submissions

Project Repository

• Public GitHub/GitLab repository with your code

• Must include a /prompts directory containing your PDD prompts

• README with setup instructions

Demo Video (3-5 minutes)

• Show your application in action

• Explain how you used PDD methodology

• Highlight Toolhouse.ai integration (if used) 

Pitch Deck (5-10 slides)

• Problem statement and market opportunity

• Solution overview and technical approach

• Business model and go-to-market strategy

• Team background

PDD Artifacts

• At least one module with:

• Prompt file (.prompt)

• Generated code

• Usage example

• Test file

• Demonstrate the regeneration workflow

Submission Checklist

Bonus Points

• Multiple modules using PDD methodology

• Evidence of test accumulation (tests preserved across regenerations)

• Integration with PDD Cloud for grounding/few-shot examples

• Clear documentation of the prompt-to-code relationship