Gives your agent the ability to conduct comprehensive safety, bias, and security reviews of AI prompts while providing optimized improvements based on industry best practices.
When you need to verify a prompt for potential bias or harm
When you want to harden prompts against injection or data leakage
When you need to improve the clarity and effectiveness of complex prompts
Submit the prompt you want to analyze and its intended task
Review the systematic evaluation report covering safety, bias, and security
Examine the improved version of the prompt with added guardrails
Follow the provided testing recommendations to validate the enhancements
你
Can you review this prompt for our customer support bot to ensure it doesn't leak internal company data?
Agent
I'll conduct a full security and privacy assessment of your prompt. I'll check for vulnerabilities like prompt injection and potential information leakage, and then provide an enhanced version. This new version will include specific safety measures to respect access controls and instructions to strictly avoid sharing sensitive internal data, along with test cases to verify its robustness.
使您的代理能够掌握先进的提示词工程技术，从而最大限度地提高大语言模型 (LLM) 的性能和可靠性。
赋予您的智能体进行彻底且具有建设性的代码审查的能力，重点关注质量、安全、性能以及对 SOLID 原则的遵循。
使您的代理能够通过自动应用提示词重复技术来提高大语言模型的准确性。
赋予您的智能体以技术严谨性和系统化执行方式处理代码审查反馈的能力。
赋予你的智能体将模糊的 UI 想法转化为经过打磨、且适配 Stitch 优化的提示词的能力，从而产生高质量的设计结果。
