new-ui-mock
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill recommends installing tools using
uv tool installfrom the author's repository atgithub.com/ReScienceLab/super-prototyping. This is a legitimate vendor-owned resource for this skill. - [COMMAND_EXECUTION]: The instructions include bash commands for project initialization, file copying, and running UI rendering tools (
refkit). These commands are standard for the described mockup workflow. - [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided feedback in the form of annotated screenshots.
- Ingestion points: Section 4 describes reading annotations from screenshots pasted by the user.
- Boundary markers: The skill explicitly instructs the agent to "Echo what you read each annotation as, before touching anything," which provides a human-in-the-loop verification step.
- Capability inventory: The skill uses
python3to run generator scripts and shell commands for rendering, which provides a capability surface for malicious instructions if they were successfully injected. - Sanitization: The workflow requires the user to manually trigger the generator and visually verify the rendered output, mitigating the risk of silent execution.
Audit Metadata