Pass
Audited by Gen Agent Trust Hub on Jul 2, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill documentation suggests installing several third-party libraries for PDF processing, such as pypdf, pdfplumber, reportlab, pytesseract, pdf2image, pypdfium2, pdf-lib, and pdfjs-dist. These are established, widely-used libraries for the intended tasks.\n- [COMMAND_EXECUTION]: Instructions describe how to use command-line tools such as pdftotext, qpdf, pdftk, and pdftoppm to manipulate files. These commands are typical for document processing workflows.\n- [DATA_EXFILTRATION]: The skill reads and processes PDF files, which could contain sensitive user data. However, there is no evidence of network requests or exfiltration to external domains.\n- [PROMPT_INJECTION]: The skill processes untrusted PDF data through text extraction and OCR, creating a surface for indirect prompt injection.\n
- Ingestion points: PDF files are read using libraries like pypdf, pdfplumber, and pytesseract.\n
- Boundary markers: None implemented in the provided extraction scripts.\n
- Capability inventory: The skill has the ability to write files and execute local shell commands for PDF processing.\n
- Sanitization: Extracted text is not sanitized before being returned to the agent.
Audit Metadata