image
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides example shell commands for image optimization using common utilities such as
cwebp,mogrify, andjpegoptim. These commands are intended to operate on local image files and are consistent with the skill's purpose. - [EXTERNAL_DOWNLOADS]: The skill references multiple external APIs and platforms for AI image generation, including Google Gemini, OpenAI, Vercel, and Ideogram. All external links target official domains of well-known and trusted providers.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a surface for processing untrusted data by reading local context files (
product-marketing.md) and fetching image metadata from websites viacurl. Ingestion points: project context files and remote website HTML. Boundary markers: absent. Capability inventory:cwebp,mogrify,jpegoptim, andcurl. Sanitization: absent. These elements are part of a standard workflow for marketing automation and context gathering. - [SAFE]: No obfuscation, persistence mechanisms, or credential exposures were detected.
Audit Metadata