skills/ntaksh42/agents/pdf-processor/Gen Agent Trust Hub

pdf-processor

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • Indirect Prompt Injection (LOW): The skill extracts text from external PDF files which serves as a potential ingestion point for untrusted data. Evidence: (1) Ingestion point: reader.pages.extract_text() in SKILL.md. (2) Boundary markers: None present in code samples. (3) Capability inventory: File system read/write and PDF manipulation. (4) Sanitization: No sanitization or validation of extracted text is performed.
  • External Dependencies (SAFE): The skill references established libraries such as PyPDF2, pdf-lib, and weasyprint from official registries.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:32 PM