direccion-de-arte
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists of instructional markdown files providing templates and methodologies for image prompt generation. It does not contain executable code, scripts, or network-enabled commands.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from user image descriptions and brand configuration files (marca.json) to construct prompts for image generation tools. While capability exists to trigger image generation and editing via external tools, the structured block-based generation process acts as a constraint. No explicit sanitization or boundary markers are used for the ingested data.
- [COMMAND_EXECUTION]: The skill references intentions for generating and editing images that are intended to be interpreted by a separate tool provider. There is no evidence of arbitrary command execution or user-supplied input flowing into shell commands.
Audit Metadata