Function main

backend/agent/scripts/test_code_agent_live.py:321–476

()

def main():
    parser = argparse.ArgumentParser(description="Live test CodeAgent quality")
    parser.add_argument("--model", default="gpt-4.1", help="Model to use")
    parser.add_argument("--test-id", type=int, default=0, help="Test case ID (0=all)")
    parser.add_argument("--no-sandbox", action="store_true", help="Skip sandbox exec-fix")
    parser.add_argument("--tree-search", action="store_true", help="Enable tree search")
    parser.add_argument("--output-dir", default="test_outputs", help="Output directory")
    args = parser.parse_args()

    # Setup LLM client
    base_url = os.environ.get("OPENAI_BASE_URL", "")
    api_key = os.environ.get("OPENAI_API_KEY", "")
    if not base_url or not api_key:
        print("ERROR: Set OPENAI_BASE_URL and OPENAI_API_KEY environment variables")
        sys.exit(1)

    llm_config = LLMConfig(
        base_url=base_url,
        api_key=api_key,
        primary_model=args.model,
        fallback_models=[],
        max_tokens=16384,
        temperature=0.7,
        timeout_sec=300,
    )
    llm = LLMClient(llm_config)

    # Quick connectivity test
    print(f"Testing LLM connectivity ({args.model})... ", end="", flush=True)
    ok, msg = llm.preflight()
    if not ok:
        print(f"FAILED: {msg}")
        sys.exit(1)
    print("OK")

    pm = PromptManager()

    # Select test cases
    if args.test_id > 0:
        if args.test_id not in TEST_CASES:
            print(f"ERROR: Unknown test ID {args.test_id}. Available: {list(TEST_CASES.keys())}")
            sys.exit(1)
        cases = {args.test_id: TEST_CASES[args.test_id]}
    else:
        cases = TEST_CASES

    # Output directory
    output_dir = Path(args.output_dir)
    output_dir.mkdir(parents=True, exist_ok=True)

    all_reports = []

    for test_id, tc in cases.items():
        print(f"\n{'='*60}")
        print(f"Test {test_id}: {tc['name']}")
        print(f"{'='*60}")

        stage_dir = output_dir / f"test_{test_id}"