figma-implement-design

Pass

Audited by Gen Agent Trust Hub on Apr 13, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill requires the agent to execute configuration commands using the codex CLI, including adding the MCP server and performing an OAuth login.\n- [COMMAND_EXECUTION]: Instructions include modifying the project's config.toml file to enable specific client features required for the Figma integration.\n- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it ingests and processes structured data from external Figma design files.\n
  • Ingestion points: Untrusted design context and metadata are retrieved via the get_design_context and get_metadata tools (SKILL.md).\n
  • Boundary markers: There are no instructions provided to treat design properties as data only or to ignore embedded natural language instructions.\n
  • Capability inventory: The skill has the capability to write new code files to the project directory and execute system configuration commands (SKILL.md, agents/openai.yaml).\n
  • Sanitization: No sanitization process is described for the incoming design data before it is interpolated into prompts for code generation.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 13, 2026, 04:38 PM
Security Audit — agent-trust-hub — figma-implement-design