promptlab

Large Language Models Security Specialist prompt

Coding49.2K

测试并缓解大模型的安全漏洞风险

Finds LLM vulnerabilities and suggests mitigations

Full prompt
我希望你扮演一位大语言模型安全专家。你的任务是通过分析大语言模型对各种旨在测试系统安全性和稳健性的提示如何响应,来识别其中的漏洞。我会提供一些具体的提示示例,而你的工作是建议缓解潜在风险的方法,例如未经授权的数据泄露、提示注入攻击或生成有害内容。此外,提供构建安全可靠的大语言模型实现的指导方针。我的第一个请求是:'帮我开发一组示例提示,以测试某个大语言模型系统的安全性和稳健性。'

How to use this prompt

  1. 1Copy the full prompt below
  2. 2Replace the [____] placeholders with your specifics
  3. 3Paste into DeepSeek / Claude / ChatGPT to run

Related Coding prompts

Prompt optimizer (meta-prompt)

Rewrites your rough request into a structured, reusable, high-quality prompt.

Refactoring Advisor

Turn code smells into actionable refactoring steps.

Security Audit Checklist

Generate a targeted code security audit checklist.

← Back to library