inference.py

import os
import json
import requests
from openai import OpenAI
from dotenv import load_dotenv

# Load environment variables from .env file
load_dotenv()

# Required environment variables for OpenEnv baseline agents
# These will be provided by the hackathon evaluation environment
API_BASE_URL = os.getenv("API_BASE_URL", "https://api.openai.com/v1")
MODEL_NAME = os.getenv("MODEL_NAME", "gpt-4o")
HF_TOKEN = os.getenv("HF_TOKEN", "")

# Hugging Face Space URL (Replace with your actual URL after deployment)
# For local testing, use http://localhost:8000
ENV_URL = os.getenv("ENV_URL", "http://localhost:8000")

client = OpenAI(base_url=API_BASE_URL, api_key=HF_TOKEN)

def call_llm(contract_code, task_description):
    """Ask the LLM to audit a smart contract"""
    print(f"Auditing contract with {MODEL_NAME}...")
    
    response = client.chat.completions.create(
        model=MODEL_NAME,
        messages=[
            {"role": "system",
             "content": "You are a specialized Solidity security auditor. Your goal is to find vulnerabilities in smart contracts."},
            {"role": "user",
             "content": f"Audit this contract:\n\n{contract_code}\n\n"
                f"Task: {task_description}\n\n"
                "Return valid JSON ONLY in this format: {\"analysis\": \"...\", "
                "\"vulnerabilities\": [{\"type\": \"...\", "
                "\"line\": N, \"severity\": \"...\", "
                "\"description\": \"...\"}], "
                "\"suggested_fixes\": [\"...\"]}"
            }
        ],
        temperature=0.1,
        response_format={ "type": "json_object" } # Ensure JSON output
    )
    
    content = response.choices[0].message.content
    try:
        return json.loads(content)
    except json.JSONDecodeError:
        print(f"Failed to parse LLM response: {content}")
        return {
            "analysis": "error",
            "vulnerabilities": [],
            "suggested_fixes": []
        }

def run_task(task_id):
    """Run one task against the environment"""
    print(f"\n--- Starting Task: {task_id} ---")
    
    # 1. Reset environment for this task
    r = requests.post(f"{ENV_URL}/reset",
                      json={"task_id": task_id})
    if r.status_code != 200:
        print(f"Error resetting environment: {r.text}")
        return 0.0
        
    obs = r.json()
    
    # 2. Call LLM to audit the contract
    result = call_llm(obs["contract_code"],
                      obs["task_description"])
    
    # 3. Step the environment with the agent's findings
    r = requests.post(f"{ENV_URL}/step", json={
        "analysis": result.get("analysis", ""),
        "vulnerabilities": result.get("vulnerabilities", []),
        "suggested_fixes": result.get("suggested_fixes", [])
    })
    
    if r.status_code != 200:
        print(f"Error stepping environment: {r.text}")
        return 0.0
        
    step_obs = r.json()
    reward = step_obs.get("reward", 0.0)
    feedback = step_obs.get("feedback", "No feedback")
    
    print(f"Task: {task_id} | Score: {reward:.2f} | Feedback: {feedback}")
    return reward

if __name__ == "__main__":
    # Ensure environment is running
    try:
        requests.get(ENV_URL)
    except requests.exceptions.ConnectionError:
        print(f"Warning: Could not connect to {ENV_URL}. Make sure your environment server is running.")

    tasks = ["easy_reentrancy", "medium_multi_vuln", "hard_full_audit"]
    total_score = 0
    
    for task_id in tasks:
        score = run_task(task_id)
        total_score += score
        
    print(f"\nFinal Total Score: {total_score:.2f} / {len(tasks)}.00")