From Prompty to Code

In this section, we’ll learn how to convert Prompty assets to Python code and build your first production-ready application.

Prerequisites

Before we begin, ensure you have:

• Python 3.10 or higher installed
• Visual Studio Code with the Prompty extension
• A completed Prompty asset from the previous section
• Access to Azure OpenAI or another supported model provider

Installation

First, install the Prompty Python package with the appropriate extras:

# For Azure OpenAI
pip install "prompty[azure]"

# For OpenAI
pip install "prompty[openai]"

# For GitHub Models and serverless
pip install "prompty[serverless]"

# Install all providers
pip install "prompty[azure,openai,serverless]"

Generate Code from Prompty Asset

The VS Code extension can automatically generate Python code from your Prompty assets:

1. Open your Prompty file in VS Code
2. Right-click on the file in the Explorer
3. Select “Add Code” → “Add Prompty Code”
4. Choose your preferred template (basic, with tracing, async, etc.)

This generates a Python file with the necessary imports and execution logic.

Generated Code Example

Here’s what the generated code looks like:

import prompty
import prompty.azure
from prompty.tracer import trace, Tracer, console_tracer, PromptyTracer

# Configure tracing (optional)
Tracer.add("console", console_tracer)
json_tracer = PromptyTracer(output_dir="./traces")
Tracer.add("json", json_tracer.tracer)

@trace
def run(question: str) -> str:
    """Execute the Shakespeare prompt with the given question"""

    # Execute the prompty file
    response = prompty.execute(
        "shakespeare.prompty",
        inputs={"question": question}
    )

    return response

if __name__ == "__main__":
    # Example usage
    result = run("What is the meaning of life?")
    print(result)

Environment Configuration

Set up Environment Variables

Create a .env file in your project directory:

.env

AZURE_OPENAI_ENDPOINT=https://your-resource.openai.azure.com/
AZURE_OPENAI_API_KEY=your-api-key
AZURE_OPENAI_DEPLOYMENT=gpt-35-turbo
AZURE_OPENAI_API_VERSION=2024-10-21

Load Environment Variables

Update your Python code to load environment variables:

import os
from dotenv import load_dotenv
import prompty
import prompty.azure

# Load environment variables
load_dotenv()

def run(question: str) -> str:
    """Execute the Shakespeare prompt with environment configuration"""

    response = prompty.execute(
        "shakespeare.prompty",
        inputs={"question": question}
    )

    return response

if __name__ == "__main__":
    # Verify environment variables are loaded
    if not os.getenv("AZURE_OPENAI_ENDPOINT"):
        print("Error: AZURE_OPENAI_ENDPOINT not found in environment")
        exit(1)

    result = run("To be or not to be?")
    print(result)

Building a Simple Application

Let’s create a more comprehensive application:

app.py

import os
import sys
from dotenv import load_dotenv
import prompty
import prompty.azure
from prompty.tracer import trace, Tracer, console_tracer

# Load environment variables
load_dotenv()

# Configure tracing for development
Tracer.add("console", console_tracer)

The Prompty runtime supports additional runtimes, including frameworks such as [LangChain](/tutorials/using-langchain/), and [Semantic Kernel](/tutorials/using-semantic-kernel/). In the tutorials section, we will cover how to generate code from Prompty assets using these runtimes. (coming soon)

def main():
    """Main application loop"""
    print("🎭 Welcome to the Shakespeare Chat App!")
    print("Type 'quit' to exit, 'help' for commands")

    while True:
        user_input = input("\n🤔 Ask Shakespeare: ").strip()

        if user_input.lower() in ['quit', 'exit', 'q']:
            print("👋 Farewell!")
            break
        elif user_input.lower() == 'help':
            print("Commands:")
            print("  help - Show this help")
            print("  quit - Exit the application")
            print("  Just type any question to chat with Shakespeare!")
            continue
        elif not user_input:
            continue

        print("🎭 Shakespeare is thinking...")
        response = chat_with_shakespeare(user_input)

        if response:
            print(f"🎭 Shakespeare: {response}")
        else:
            print("😔 Shakespeare seems to be having trouble responding.")

if __name__ == "__main__":
    # Verify required environment variables
    required_vars = [
        "AZURE_OPENAI_ENDPOINT",
        "AZURE_OPENAI_API_KEY",
        "AZURE_OPENAI_DEPLOYMENT"
    ]

    missing_vars = [var for var in required_vars if not os.getenv(var)]

    if missing_vars:
        print(f"❌ Missing required environment variables: {', '.join(missing_vars)}")
        print("Please check your .env file and try again.")
        sys.exit(1)

    main()

Running Your Application

Command Line Execution

Run your application from the command line:

# Basic execution
python app.py

# With verbose tracing
PROMPTY_TRACE_LEVEL=DEBUG python app.py

In VS Code

1. Open the integrated terminal (Ctrl+`)
2. Navigate to your project directory
3. Run the application: python app.py

Error Handling and Debugging

Add Robust Error Handling

import logging
from typing import Optional

# Configure logging
logging.basicConfig(
    level=logging.INFO,
    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s'
)
logger = logging.getLogger(__name__)

def safe_execute_prompt(prompt_path: str, inputs: dict) -> Optional[str]:
    """Safely execute a prompt with comprehensive error handling"""
    try:
        logger.info(f"Executing prompt: {prompt_path}")

        response = prompty.execute(prompt_path, inputs=inputs)

        logger.info("Prompt executed successfully")
        return response

    except FileNotFoundError:
        logger.error(f"Prompt file not found: {prompt_path}")
        return None
    except ConnectionError:
        logger.error("Failed to connect to AI service")
        return None
    except Exception as e:
        logger.error(f"Unexpected error: {e}")
        return None

# Usage
response = safe_execute_prompt("shakespeare.prompty", {"question": "Hello"})
if response:
    print(response)
else:
    print("Failed to get response")

Next Steps

Now that you have a working application, explore these advanced topics:

1. Add Multiple Prompts

Create different prompty files for different use cases and route requests appropriately.

2. Add Web Interface

Create a simple web API with FastAPI to serve your prompts over HTTP.

3. Explore Advanced Features

• Observability & Tracing - Monitor your application
• Performance Optimization - Scale your application
• Configuration - Manage different environments
• CLI Usage - Use the command-line interface

Common Issues and Solutions

Issue: ImportError

# Error: No module named 'prompty'
pip install "prompty[azure]"

# Error: No module named 'dotenv'
pip install python-dotenv

Issue: Authentication Errors

# Verify your credentials
import os
print(f"Endpoint: {os.getenv('AZURE_OPENAI_ENDPOINT')}")
print(f"API Key: {'***' if os.getenv('AZURE_OPENAI_API_KEY') else 'Not set'}")
print(f"Deployment: {os.getenv('AZURE_OPENAI_DEPLOYMENT')}")

Issue: File Not Found

import os
prompt_path = "shakespeare.prompty"
if os.path.exists(prompt_path):
    print(f"✅ Found prompt file: {prompt_path}")
else:
    print(f"❌ Prompt file not found: {prompt_path}")
    print(f"Current directory: {os.getcwd()}")
    print("Files in directory:", os.listdir("."))

Congratulations! You’ve successfully converted your Prompty asset to code and built your first application. 🎉

7. Additional supported runtimes

The Prompty runtime supports additional runtimes, including frameworks such as LangChain, and Semantic Kernel. In the tutorials section, we will cover how to generate code from Prompty assets using these runtimes. (coming soon)

---

Want to Contribute To the Project? - Updated Guidance Coming Soon.