image-superres-enhance

Pass

Audited by Gen Agent Trust Hub on Jun 21, 2026

Risk Level: SAFE
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill accesses API credentials located at ~/.meitu/credentials.json for service authentication. This is standard behavior for the vendor's provided tool.
  • [COMMAND_EXECUTION]: The skill invokes the meitu CLI to perform image restoration and enhancement tasks.
  • [PROMPT_INJECTION]: The skill processes user-supplied image URLs and agent-generated content descriptions within its execution logic. Ingestion points: image_url and prompt parameters; Boundary markers: Double quotes used in shell command template; Capability inventory: Execution of meitu CLI and file writes to ./output/; Sanitization: Not explicitly defined in instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 21, 2026, 12:30 AM
Security Audit — agent-trust-hub — image-superres-enhance