Pass
Audited by Gen Agent Trust Hub on May 18, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute local Python scripts and system utilities such as
qpdf,pdftotext, andmagickto perform PDF and image processing. These operations are core functionality for the skill's purpose and do not involve untrusted remote sources. - [SAFE]: A detailed review of the provided scripts and instructions found no evidence of prompt injection, data exfiltration, or obfuscated code. The skill's implementation follows standard practices for document processing.
- [PROMPT_INJECTION]: While the skill processes untrusted PDF documents (representing an indirect prompt injection surface), it does so within a defined scope focused on data extraction.
- Ingestion points: PDF files read via
pypdf,pdfplumber, andpdftotext. - Boundary markers: None explicitly defined in the helper scripts.
- Capability inventory: Local file access and execution of PDF processing binaries.
- Sanitization: Relies on the parsing logic of standard libraries; no explicit content filtering is present.
Audit Metadata