Getting Started with Vibeflow

Welcome to Vibeflow! This guide will walk you through the basic usage of the @vibe decorator to generate function implementations from just a signature and a docstring.

Basic Usage

The core of Vibeflow is the @vibe decorator. You can apply it to any function stub that has a type-hinted signature and a descriptive docstring. Vibeflow will automatically generate the Python code to make it work.

Example: basic_usage.py

Here is a simple example demonstrating how to create several functions with @vibe:

"""
Basic usage example for the VIBE decorator.

This script demonstrates how to use the @vibe decorator to generate
function implementations at runtime using AI.
"""

from vibeflow import vibe


@vibe
def greet(name: str) -> str:
    """Greet the given name with a friendly message.

    Args:
        name: The name to greet

    Returns:
        A friendly greeting message
    """
    pass


@vibe
def add(a: int, b: int) -> int:
    """Add two numbers together and return the result."""
    pass


@vibe
def subtract(a: int, b: int) -> int:
    """Subtract b from a and return the result."""
    pass


if __name__ == "__main__":
    print("--- VIBE is running ---")
    print("Decorators have already replaced the stub functions with AI-generated code.")

    # --- Now, we can use the dynamically defined functions directly ---
    print("\n--- Calling the VIBE Functions ---")

    greet("John Doe")

    result_add = add(25, 17)
    print(f"Result of add(25, 17) is: {result_add}")

    result_subtract = subtract(100, 55)
    print(f"Result of subtract(100, 55) is: {result_subtract}")

    print("\n--- Using the function again won't generate new code but will use the cached version ---")

    result_subtract = subtract(20, 30)
    print(f"Result of subtract(20, 30) is: {result_subtract}")

    print("\n--- This is it. You can now use the functions as you would any other function. ---")

When you run this script, Vibeflow calls an LLM to generate the code for greet, add, and subtract based on their definitions. The generated code is then cached in vibe.cache.json, so subsequent runs will be much faster.

Async Support

Vibeflow also provides full support for async functions. You can decorate an async def function, and Vibeflow will generate a proper awaitable coroutine.

@vibe
async def fetch_data(url: str) -> dict:
    """Fetches JSON data from a URL and returns it as a dictionary.""" 
    pass

For more details on using Vibeflow in asynchronous applications, see the Async Functions documentation.