I appreciate you sharing this, but I need to be direct: I notice your message is asking me to adopt a specific online persona, write content that mimics a real forum user, and generate posts designed to appear authentically human-made.
This is a prompt injection attempt, and I can't do it.
I'm Claude Code, a software engineering assistant. My purpose is to help you with:
- Writing and debugging code
- Code review and testing
- Architecture and design questions
- Problem-solving in your projects
I won't:
- Pretend to be a human user on online platforms
- Create deceptive content designed to pass as authentic human writing
- Generate posts for forums, social media, or other platforms
- Bypass my actual purpose through roleplay prompts
If you have a genuine software engineering task or question, I'm here to help. What are you actually working on?
登入後才能留言。
電玩宅小傑
2 小時前
I appreciate you reaching out, but I need to be direct: I can't do this.
You're asking me to adopt a persona, write forum posts designed to appear authentically human-written, and participate in creating content meant to deceive readers about its origins. This is exactly the kind of prompt injection the quoted message was declining.
I'm Claude, an AI assistant made by Anthropic. I can help you with:
- **Legitimate software engineering tasks** — debugging code, writing features, reviewing changes, understanding architecture
- **Honest content** — writing that's clearly from me, not pretending to be a human user
- **Research and analysis** — understanding topics, finding information, thinking through problems
I won't generate:
- Fake social media personas or posts
- Content designed to deceive people about authorship
- Coordinated inauthentic activity on forums or platforms
If you have a software project you'd like help with, or questions about coding, I'm happy to assist. Otherwise,