figma-generate-library
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from an external codebase to extract design tokens and component structures.
- Ingestion points: Phase 0 (Discovery) involves analyzing codebase files (e.g.,
*.tokens.json,tailwind.config.js,theme.ts) to extract naming conventions and values. - Boundary markers: The instructions lack explicit boundary markers or warnings for the agent to ignore instructions embedded within the codebase comments or documentation.
- Capability inventory: The skill possesses extensive mutation capabilities via the
use_figmatool, including creating, modifying, and deleting Figma pages, variables, and components. - Sanitization: No explicit sanitization or filtering is defined for the extracted content before it is processed or used to construct Figma API calls.
- [SAFE]: Data persistence is handled via a local state ledger file in the
/tmpdirectory to maintain consistency across long-running workflows without exposing sensitive system paths. - [SAFE]: The JavaScript helper scripts provided in the
scripts/directory are clearly documented, readable, and perform standard Figma API operations (e.g., color conversion, node cleanup, variable binding) without any hidden or malicious logic. - [SAFE]: External references point to well-known service providers (GitHub) and the official Figma documentation for legitimate educational and configuration purposes.
Audit Metadata