using-figma

Pass

Audited by Gen Agent Trust Hub on Jun 14, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill implements a robust least-privilege model by using the allowed-tools frontmatter to restrict the agent's environment to read-only Figma tools, explicitly excluding mutation capabilities.- [PROMPT_INJECTION]: The skill processes untrusted external data in the form of design metadata, layer names, and strings fetched from Figma, creating an indirect prompt injection surface.
  • Ingestion points: Data is ingested via the mcp__figma__get_design_context, mcp__figma__get_metadata, and mcp__figma__get_variable_defs tools in SKILL.md.
  • Boundary markers: No explicit delimiters or 'ignore instructions' markers are defined for the design context when it is interpolated into the prompt.
  • Capability inventory: The skill's capabilities are strictly limited to the referenced read-only tools and it does not possess file-write, network exfiltration, or arbitrary execution permissions.
  • Sanitization: The instructions do not specify any sanitization, escaping, or filtering for content retrieved from Figma layers or text nodes.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 14, 2026, 08:39 AM
Security Audit — agent-trust-hub — using-figma