Detect invisible instructions, white-colored text, and AI prompt injections secretly embedded inside documents — instantly, in your browser.
or click to browse from your computer
No signup. No server uploads. No complexity. Upload and reveal in seconds.
Drag and drop or select any PDF or DOCX file. Everything runs inside your browser — no server involved.
All text including white, tiny, and off-page content is extracted and displayed in orange so nothing stays invisible.
Our threat engine scans for 20+ injection patterns and flags dangerous AI manipulation instructions automatically.
Full text layer extraction from both major document formats with metadata inspection.
Reveals white text, zero-size fonts, off-page content, and zero-width unicode characters.
Forces all extracted text to orange so previously invisible white text becomes clearly visible.
Files are processed in your browser only. Nothing is ever sent to or stored on any server.
Results appear in seconds using native browser PDF and DOCX parsing APIs.
Copy all extracted text or download a full threat report as a TXT file.
Detects persona injections, jailbreaks, LLM format tokens, base64 obfuscation, and more.
Search through all extracted blocks and filter to show only suspicious content instantly.
Verify academic submissions contain no hidden AI manipulation instructions.
Test document-based AI pipelines against real-world prompt injection attacks.
Audit PDF files for embedded attack payloads and obfuscated instructions.
Ensure contract documents contain no hidden or manipulated content.
Inspect source documents for hidden editorial instructions or tampering signs.
Files are never uploaded to any server
All processing happens inside your browser
No account or login required ever
No tracking of uploaded documents
Files cleared from memory after analysis
100% free with no hidden costs
Your document has been successfully analyzed. Review the results above or start a new scan below.