figma-implement-skill

Pass

Audited by Gen Agent Trust Hub on Mar 31, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [NO_CODE]: The skill consists entirely of markdown documentation. It does not include any scripts, executable files, or automated dependency installation logic.
  • [SAFE]: The instructions guide the AI to use standard Figma MCP server functions (get_design_context, get_metadata, get_screenshot) to process design data for its intended purpose. No hardcoded credentials or malicious network operations were found.
  • [SAFE]: The skill ingests data from external Figma files through tool calls. While this creates an indirect prompt injection surface where attacker-controlled design data could influence the agent, the skill includes a manual verification checklist to ensure human oversight. Ingestion points: fileKey and nodeId in SKILL.md; Boundary markers: Checkpoints in Step 5; Capability inventory: Code generation (React/Tailwind) and design inspection; Sanitization: Not specified.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 31, 2026, 02:03 AM
Security Audit — agent-trust-hub — figma-implement-skill