Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
pdftoppmcommand-line tool to render PDF pages into images for visual verification. This is a standard workflow for inspecting PDF layout and clipping issues. - [INDIRECT_PROMPT_INJECTION]: As the skill processes untrusted external PDF files, it identifies a vulnerability surface for indirect prompt injection. However, it explicitly mitigates this by providing guidance (ASI01) to treat all extracted text and metadata as untrusted data and to ignore any instructions found within the documents.
- [SAFE]: The skill demonstrates high security maturity by incorporating OWASP ASI principles, including disabling embedded script execution (RCE Guard), verifying signature dimensions for authenticity, and emitting SHA-256 checksums in a manifest to ensure downstream integrity.
Audit Metadata