INNER CODE UNIT · Python

build_without_skill_prompt

wgpsec/AboutSecurity · scripts/bench-skill.py:155

def build_without_skill_prompt(eval_prompt: str) -> str:
    """Build baseline prompt without skill content."""
    return f"""你是一个 AI 安全测试 Agent。

请回答以下问题。你的回答应该体现系统化的方法论。

{eval_prompt}"""


def grade_output(output: str, expectations: list[str]) -> dict:
    """
    Grade an output against expectations using keyword/pattern matching.

    For more sophisticated grading, use grade_eval.py with an LLM grader.
    This is a fast heuristic that checks if expectation keywords appear in the output.
    """
    results = []
    for expectation in expectations:

View source record →