product-spec-pdf-parser

Pass

Audited by Gen Agent Trust Hub on May 11, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface (Category 8).
  • Ingestion points: Untrusted text data is extracted from user-provided PDF files via the PyMuPDF library as described in SKILL.md (Step 2).
  • Boundary markers: The workflow lacks explicit delimiters or system instructions to ignore embedded commands within the extracted text before it is processed by the model in Step 3.
  • Capability inventory: The skill has access to powerful capabilities including Bash for command execution, Write for file system modification, and Google Sheets update tools (mcp__google__sheets_values_update).
  • Sanitization: No validation or sanitization of the extracted PDF content is performed prior to interpolation into the model's context.
  • [COMMAND_EXECUTION]: The skill involves dynamic script generation and execution (Category 10).
  • Evidence: SKILL.md (Step 2) instructs the agent to write and execute a Python script using Bash to perform PDF text extraction. While the script is a simple template using the well-known PyMuPDF library, the pattern of executing generated code at runtime is a risk factor.
Audit Metadata
Risk Level
SAFE
Analyzed
May 11, 2026, 11:58 AM
Security Audit — agent-trust-hub — product-spec-pdf-parser