Media
3 items
Description
Detects hidden text attacks on web pages that manipulate AI assistants like ChatGPT and Claude. SafePrompt - AI Prompt Injection Detector Protect yourself from hidden text injection attacks that manipulate AI assistants like ChatGPT, Claude, Gemini, and Perplexity. WHAT IS HIDDEN TEXT INJECTION? Attackers embed invisible text on web pages using CSS tricks (white text on white background, zero opacity, tiny fonts). When you ask an AI to analyze a page, it reads this hidden text and follows the attacker's instructions instead of helping you. HOW SAFEPROMPT PROTECTS YOU This extension automatically scans every page you visit for: • White-on-white hidden text • Zero opacity elements • Off-screen positioned content • Microscopic font sizes • Display:none elements with suspicious content • Clip-path hidden text FEATURES Real-Time Protection • Automatic scanning on every page load • Instant alerts when threats are detected • Badge counter shows number of hidden elements found Detailed Analysis • Click the extension to see exactly what was found • View the hidden text content • Understand what technique was used Privacy First • All scanning happens locally in your browser • No data sent to external servers • No tracking or analytics WHO NEEDS THIS? • Anyone using AI assistants for research • Developers building AI-powered applications • Security researchers studying prompt injection • Journalists and fact-checkers ABOUT SAFEPROMPT SafePrompt is an AI security company focused on protecting applications from prompt injection attacks. Learn more: https://safeprompt.dev Privacy Policy: https://safeprompt.dev/privacy