product-shots-ad-creative

Pass

Audited by Gen Agent Trust Hub on Jun 21, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: Prompt Security. The skill contains no instructions designed to override agent behavior or bypass safety guidelines. It features an automated sanitation process to prevent platform-specific keywords from leaking into image generation prompts, which prevents the creation of deceptive visual artifacts.
  • [SAFE]: Data Security. Analysis of the skill body and reference files found no instances of hardcoded credentials, access to sensitive file paths, or unauthorized network operations.
  • [SAFE]: Mitigation of Indirect Prompt Injection. The skill manages untrusted data from ingestion points (user_request and brand_kit) but effectively mitigates risk through clear boundary markers (quoting user content) and a comprehensive sanitation process (banned-words filtering). The capability inventory is limited to peer skill calls (product-shots, product-shots-image-gen) with no access to dangerous system commands, file writes, or external network requests.
  • [SAFE]: Execution Safety. All operational logic is provided as descriptive pseudocode. No instances of remote code execution, privilege escalation, or dynamic code generation were identified.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 21, 2026, 10:51 AM
Security Audit — agent-trust-hub — product-shots-ad-creative