image_analyzer.py

import os
import openai
from smolagents import Tool

openai.api_key = os.getenv("OPENAI_API_KEY")

class ImageAnalyzer(Tool):
    name = "image_analyzer"
    description = "Analyze the given image and describe or reason about its contents."
    inputs = {
        "image_path": {
            "type": "string",
            "description": "Path to the image file (e.g., a chessboard image)."
        },
        "question": {
            "type": "string",
            "description": "The question to answer about the image (e.g., best chess move)."
        }
    }
    output_type = "string"

    def forward(self, image_path: str, question: str) -> str:
        with open(image_path, "rb") as image_file:
            response = openai.chat.completions.create(
                model="gpt-4-vision-preview",
                messages=[
                    {"role": "user", "content": [
                        {"type": "text", "text": question},
                        {"type": "image_url", "image_url": {"url": "data:image/jpeg;base64," + image_file.read().encode("base64").decode()}}
                    ]}
                ],
                max_tokens=500
            )
            return response.choices[0].message.content.strip()