explore
Pass
Audited by Gen Agent Trust Hub on Nov 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read project-specific files like
DESIGN.md. This ingestion point could be leveraged for indirect prompt injection if the project files contain adversarial instructions. \n - Ingestion points:
DESIGN.md(root directory), project design systems, and user requests. \n - Boundary markers: Absent; the skill does not define specific separators for untrusted data. \n
- Capability inventory: Access to Figma/Paper MCP tools and shell command execution for rendering. \n
- Sanitization: No explicit sanitization or validation of the design brief content is required. \n- [COMMAND_EXECUTION]: The skill provides templates for shell commands to render and capture images of generated HTML designs. \n
- Evidence:
references/adapters/html.mdincludes a headless Chrome command line that uses local file paths and environment variables ($PWD). This is a standard utility for the skill's purpose.
Audit Metadata