skills/adobe/skills/stardust/Gen Agent Trust Hub

stardust

Pass

Audited by Gen Agent Trust Hub on May 14, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements a robust, multi-stage redesign workflow with strict provenance tracking and human-in-the-loop validation for all significant changes.
  • [COMMAND_EXECUTION]: Orchestrates local sub-commands and helper scripts (e.g., 'load-context.mjs') located within the skill's environment to manage project state and design metadata.
  • [REMOTE_CODE_EXECUTION]: Invokes the 'impeccable' tool via npx (targeting the npm registry) to perform design audits and generation, which is a documented core dependency of the Adobe-authored skill.
  • [EXTERNAL_DOWNLOADS]: Fetches design assets, such as fonts and media, from the source website being redesigned during the migration phase to ensure visual fidelity.
  • [PROMPT_INJECTION]: While the skill processes untrusted data from crawled websites (a surface for indirect injection), this risk is mitigated by a mandatory reasoning step where the agent must present its resolved plan and dimensions to the user before executing any commands.
Audit Metadata
Risk Level
SAFE
Analyzed
May 14, 2026, 07:42 PM
Security Audit — agent-trust-hub — stardust