18-image-optimization
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides standard instructions for optimizing web images, including quality, responsive delivery, and accessibility. No malicious patterns such as credential theft, remote code execution, or obfuscation were found.
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to inspect the existing repository, dependencies, and code conventions. This represents a data ingestion surface; however, no exploitable capabilities or malicious instruction chains were identified, and the behavior is standard for engineering tasks.
Audit Metadata