figma-use
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides detailed technical instructions and code patterns for developing Figma plugins. The provided examples and guidelines follow standard development practices and do not contain malicious patterns such as command injection, data exfiltration, or persistence mechanisms.
- [INDIRECT_PROMPT_INJECTION]: The skill provides patterns for reading and processing user-generated content from Figma files, creating a potential surface for indirect prompt injection. This is documented as a standard functional risk for skills interacting with user data. 1. Ingestion points: The agent reads node names, text content, and component metadata from the Figma canvas as shown in references/common-patterns.md and references/component-patterns.md. 2. Boundary markers: The skill does not prescribe specific delimiters or warnings for embedded instructions in canvas data. 3. Capability inventory: The skill enables the agent to create, modify, and delete nodes and variables in Figma files via the use_figma tool. 4. Sanitization: No explicit string sanitization for ingested canvas data is required by the provided patterns.
Audit Metadata