implement-from-design

Pass

Audited by Gen Agent Trust Hub on Apr 3, 2026

Risk Level: SAFE
Full Analysis
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The skill reads project source code to identify reusable components and design tokens. This is a legitimate functional requirement for maintaining design system consistency and does not involve exfiltrating sensitive credentials or personal data.
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute standard development commands such as lint and test to verify code quality. These operations are performed within the local project context using the project's existing configuration.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes data from external design platforms, which serves as an ingestion surface for untrusted content. However, this is inherent to its primary purpose of design-to-code automation.
  • Ingestion points: Design metadata and structures fetched via MCP tools (figma, sketch, mastergo, pixso, modao) or provided via user screenshots (SKILL.md).
  • Boundary markers: The skill does not explicitly define markers to ignore instructions within the design data, though it provides a structured workflow for implementation.
  • Capability inventory: The agent has the capability to read local source files, write report files to a specific directory, and execute shell commands (lint/test).
  • Sanitization: There are no explicit instructions to sanitize or validate the content of design metadata before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 3, 2026, 08:53 AM