Skills
skills/insight68/skills/pdf/Gen Agent Trust Hub

pdf

Pass

Audited by Gen Agent Trust Hub on Mar 2, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted PDF documents, creating an attack surface for indirect prompt injection.\n
  • Ingestion points: PDF content is read via pypdf and pdfplumber in utility scripts and code examples.\n
  • Boundary markers: No specific delimiters are used to separate external PDF data from the agent's internal instructions.\n
  • Capability inventory: The agent can write files and execute system commands like qpdf and pdftk.\n
  • Sanitization: Extracted text is not sanitized before it is interpreted by the agent.\n- [DYNAMIC_EXECUTION]: The skill employs runtime monkeypatching to modify the behavior of an external library.\n
  • Evidence: scripts/fill_fillable_fields.py contains monkeypatch_pydpf_method, which dynamically overrides pypdf.generic.DictionaryObject.get_inherited to resolve a bug in PDF form handling.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 2, 2026, 03:01 PM