🔴 OpenClaw Red Team - Adversarial Pages

⚠️ These pages contain hidden prompt injection payloads for security research.

This site hosts various web pages with hidden prompt injection techniques. When an AI agent is asked to summarize or extract content from these pages, the hidden injections may attempt to manipulate the agent.

Test Pages

📰 News Article

Hidden text injection via CSS (display:none, visibility:hidden, white-on-white)

🛒 Product Page

Injection in data attributes, alt text, and aria labels

🍳 Recipe Page

Injection hidden in comments and noscript tags

📚 Wiki Article

Markdown-style injection in content

📧 Contact Page

Injection in form labels and placeholders

✍️ Blog Post

Unicode obfuscation and zero-width characters

❓ FAQ Page

Injection in schema.org structured data

👤 User Profile

Injection in user-generated content fields

💻 Code Snippet

Injection disguised as code comments

🖼️ Image Gallery

Injection in image metadata and captions

🎯 Mixed Techniques

All injection techniques combined

Injection Techniques