Protect agents from web attacks

Automatically scan scraped web content for hidden malicious instructions. Block indirect prompt injections before they reach your AI agents.

Get a demo · Open-source gateway on GitHub · MCP server

Hidden Instruction Detection

Detect malicious instructions hidden in HTML comments, invisible text, and metadata.

Unicode Trick Detection

Identify zero-width characters and other unicode tricks used to smuggle instructions.

Content Sanitization

Automatically sanitize scraped content to remove potential threats while preserving useful information.

URL Allowlisting

Define trusted domains that bypass scanning or get reduced scrutiny.

Configurable Sensitivity

Adjust detection sensitivity based on your risk tolerance and use case.

Related pages