figma-mcp-integration

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The skill programmatically reads and acts on user-authored Figma document content (e.g., read_my_design, scan_text_nodes, get_annotations and related calls in SKILL.md) and uses that text/annotations to drive updates and create/delete nodes, so untrusted third-party content in Figma can influence agent actions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The installation steps explicitly execute remote code—curl -fsSL https://bun.sh/install | bash and a git clone of https://github.com/grab/cursor-talk-to-figma-mcp.git (followed by running bundled setup/socket commands)—so these URLs fetch and then execute external code required by the skill.