cds-design-to-code

Pass

Audited by Gen Agent Trust Hub on Oct 7, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests external data from Figma files via the Figma MCP server to generate code. This data, including component names, descriptions, and text layers, represents an untrusted input surface where malicious instructions could be embedded to influence the agent's code generation or command execution phases.
  • Ingestion points: Figma design context is fetched using the get_design_context and get_metadata tools based on a user-provided Figma URL (specified in SKILL.md Step 2).
  • Capability inventory: The skill is authorized to write React/React Native source code to the local filesystem and execute shell commands for type checking, linting, formatting, and visual validation (specified in SKILL.md Step 6).
  • Sanitization: There are no explicit instructions for sanitizing or escaping the content retrieved from Figma before it is used in the implementation logic.
  • [COMMAND_EXECUTION]: The skill is designed to discover and run local project scripts and browser automation tools to ensure implementation quality.
  • Evidence: SKILL.md Step 6 guides the agent to inspect package.json for validation scripts (e.g., yarn typecheck, npm run lint) and to use tools like Playwright, Cypress, or Puppeteer for visual parity checks.
  • Context: These actions are performed within the scope of a development environment and require user confirmation before installing new tools, aligning with standard developer operations.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 7, 2026, 09:11 AM