🚧
This page is under construction.
Schema Generation
NPi will parse and generate LLM-compatible function schemas from the type hints and the Google-style docstrings of the functions defined in your tool class. For example, consider the following tool class:
main.py
from npiai import FunctionTool, function
class MyTool(FunctionTool):
def __init__(self):
super().__init__(
name='my_tool',
description='test tool',
)
@function
def fibonacci(self, n: int) -> int:
"""
Get the nth Fibonacci number.
Args:
n: The index of the Fibonacci number in the sequence.
"""
if n == 0:
return 0
if n == 1:
return 1
return self.fibonacci(n - 1) + self.fibonacci(n - 2)NPi will parse the type hints and docstring and generate the following JSON schema for the fibonacci function:
{
"type": "function",
"function": {
"name": "fibonacci",
"description": "Get the nth Fibonacci number.",
"parameters": {
"type": "object",
"properties": {
"n": {
"type": "integer",
"description": "The index of the Fibonacci number in the sequence."
}
},
"required": ["n"]
}
}
}