asset-aware-web-design

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill includes two Node.js utility scripts (scripts/inventory-assets.mjs and scripts/validate-manifest.mjs) that the agent is instructed to execute. These scripts perform routine tasks such as scanning for image/video files, calculating hashes, and validating the structure of a JSON manifest file.
  • [DATA_EXPOSURE_AND_EXFILTRATION]: While the inventory-assets.mjs script reads file content to generate SHA-256 hashes, its scope is strictly limited to files with specific media extensions (e.g., .png, .jpg, .mp4, .glb) within the project directory. It does not access sensitive configuration files or perform network operations to transmit data.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes project-specific data like asset manifests and visual briefs. Although this ingestion point exists, the scripts and instructions are focused on metadata validation and design decisions, posing no significant risk of execution-based injection.
  • [SAFE]: The skill establishes robust guardrails for the agent, including mandatory authenticity rules (e.g., prohibiting the fabrication of product UI or customer logos) and a required disclosure for the use of placeholder assets, ensuring transparency and commercial safety.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 12:20 PM