discovered 03 Aug 2026
PDF-Prompt-Injection-Toolkit
→ View on GitHubThe PDF Prompt Injection Toolkit is a cybersecurity tool designed for red and blue teams to test and detect prompt injection attacks that may be concealed within PDF documents. It features two primary roles: a `pdf_injector.py` for creating hidden payloads and a `pdf_injection_detector.py` for scanning PDFs to identify such vulnerabilities. Notable detection techniques include scanning for invisible text, analyzing document metadata, and detecting off-page text, ensuring comprehensive coverage against potential injection tactics.