infographic
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to use the rsvg-convert utility to export SVG designs to high-resolution PNG files and the open command to preview the results. These operations are limited to image processing and viewing.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted text from users to extract data for the generated infographics, creating a potential surface for injection. * Ingestion points: User-provided text submitted via the /infographic command (SKILL.md). * Boundary markers: No explicit delimiters or instructions are provided to insulate the agent from potentially adversarial instructions embedded in the input text. * Capability inventory: The skill possesses capabilities for file system writes to the Downloads directory and command execution via shell utilities. * Sanitization: The skill does not define specific validation or sanitization routines for the input text before it is processed by the model for information extraction.
Audit Metadata