processing-pdfs

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill utilizes standard, reputable libraries for PDF processing, including pypdf, pdfplumber, and reportlab for Python, and pdf-lib for JavaScript. It also incorporates established system utilities like qpdf and poppler-utils. All identified dependencies are standard packages from official registries.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process content from external PDF files, which represents a potential surface for indirect prompt injection if documents contain malicious instructions.
  • Ingestion points: The skill reads text, metadata, and form fields from user-provided documents using libraries such as pdfplumber and pypdf (SKILL.md, references/forms.md).
  • Boundary markers: The instructions do not specify the use of delimiters or provide explicit warnings to the agent regarding potential instructions embedded within the processed document content.
  • Capability inventory: The skill enables the agent to write files to the filesystem, perform OCR on images, and execute shell-based PDF utilities (qpdf, pdftotext) based on document processing tasks.
  • Sanitization: There is no mention of text sanitization or validation of the extracted data before it is used in subsequent operations or agent responses.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 08:33 AM
Security Audit — agent-trust-hub — processing-pdfs