Pass
Audited by Gen Agent Trust Hub on Mar 30, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill performs file operations and PDF manipulations by executing local Python scripts located in
gateway/tools/pdf/and utilizing theqpdfsystem utility for merging and splitting tasks. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through untrusted PDF content. 1. Ingestion points: PDF files provided by the user (Phase 2-A, 2-D). 2. Boundary markers: No delimiters or instructions to ignore embedded content are specified in the prompt assembly. 3. Capability inventory: The skill can write files and execute local Python scripts via the
pdf-handleragent. 4. Sanitization: There is no evidence of sanitization or safety checks for the content extracted from PDF files before it is processed by the agent.
Audit Metadata