epd-parser

Pass

Audited by Gen Agent Trust Hub on May 11, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the Bash tool to execute a Python script that utilizes the PyMuPDF (fitz) library for extracting text from PDF documents. This is a functional requirement for its data parsing tasks.
  • [PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it ingests and processes untrusted text from external PDF files without explicit sanitization or strict boundary markers.
  • Ingestion points: Text content extracted from user-supplied EPD PDF documents in the extraction workflow.
  • Boundary markers: Absent. The skill does not use delimiters or specific instructions to isolate the untrusted PDF content from its reasoning logic.
  • Capability inventory: The skill has access to Bash, file system Write, and Google Sheets API tools, which could be misused if an injection is successful.
  • Sanitization: No validation or sanitization is performed on the raw text extracted from the PDFs before it is processed by the model.
Audit Metadata
Risk Level
SAFE
Analyzed
May 11, 2026, 11:58 AM
Security Audit — agent-trust-hub — epd-parser