draw
Pass
Audited by Gen Agent Trust Hub on Nov 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes the
sofficecommand-line utility to perform document conversions and graphics operations, which involves spawning local subprocesses with arguments derived from task requirements.\n- [INDIRECT_PROMPT_INJECTION]:\n - Ingestion points: The skill processes external file formats including ODG, SVG, and PDF, which are untrusted data sources that can be provided to the agent.\n
- Boundary markers: There are no specific instructions or delimiters provided to separate the content of processed documents from the agent's internal control instructions.\n
- Capability inventory: The skill has the capability to execute shell commands (soffice), write output files to the local filesystem, and establish local socket connections (localhost:8100) via the UNO bridge for automation.\n
- Sanitization: The documentation includes a safety section that warns about the risks of macros and external links in documents, recommending that users disable these features and isolate untrusted conversions.
Audit Metadata