grant-review

Pass

Audited by Gen Agent Trust Hub on May 12, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes shell commands via uv run and uvx to process PDF documents. It runs a Python script to convert PDF pages to PNG images and uses a CLI tool for markdown conversion.
  • [EXTERNAL_DOWNLOADS]: The skill fetches the opencite, pdf2image, and pillow packages from the Python Package Index (PyPI) at runtime to facilitate document processing.
  • [PROMPT_INJECTION]: The skill ingests untrusted PDF data, which presents a surface for indirect prompt injection (Category 8).
  • Ingestion points: User-provided grant proposals in PDF format are read and converted to text.
  • Boundary markers: No specific delimiters or instructions are used to distinguish processed content from system instructions.
  • Capability inventory: The skill can execute shell commands, perform file system operations, and download external code.
  • Sanitization: No validation or sanitization is mentioned for the text extracted from the proposals before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
May 12, 2026, 01:35 PM
Security Audit — agent-trust-hub — grant-review