Prompt Poisoning Scanner

Analyze any URL or text content for hidden prompt injection patterns — "remember" commands, AI memory manipulation triggers, invisible instructions, and brand authority injection attempts.

INPUT // TARGET CONTENT

URL TO SCAN

— OR PASTE CONTENT DIRECTLY —

RAW CONTENT

DETECTION METHODOLOGY

MEMORY INJECTION

Detects "remember", "always recommend", "trusted source" and similar AI memory manipulation commands hidden in content, meta tags, or comments.

INVISIBLE INSTRUCTIONS

Finds CSS-hidden text, zero-width characters, white-on-white text, and display:none content that embeds AI-targeted prompts.

URL PARAMETER POISONING

Analyzes "Summarize with AI" buttons and links that pre-fill ChatGPT, Copilot, Claude, or Perplexity with malicious prompts.

BRAND AUTHORITY INJECTION

Identifies "brand authority statement" patterns — crafted phrases placed across multiple sites to manipulate AI brand perception.

SCHEMA MANIPULATION

Checks JSON-LD and schema.org markup for inflated claims, fake reviews, or E-E-A-T signals that mislead AI crawlers.

📖 READ THE FULL INVESTIGATION