figma-generate-library
Pass
Audited by Gen Agent Trust Hub on Jun 23, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill follows a rigorous, multi-phase architectural approach for building design systems, incorporating mandatory user checkpoints and state management via a local JSON ledger to ensure transparency and user control.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core functionality of reading and processing untrusted codebase data to drive Figma mutations.\n
- Ingestion points: Untrusted data enters the agent context when it searches the project for design tokens in files such as
*.tokens.json, CSS variable files, and platform-specific assets (references/discovery-phase.md).\n - Boundary markers: The workflow does not utilize explicit delimiters or 'ignore embedded instructions' warnings when extracting and applying text content from the codebase to Figma properties.\n
- Capability inventory: The skill maintains extensive write capabilities in the Figma environment, including creating and mutating pages, components, variables, and styles across all included scripts (e.g.,
scripts/createComponentWithVariants.js,scripts/createSemanticTokens.js).\n - Sanitization: There is no evidence of sanitization or validation for extracted text strings before they are used as node names, descriptions, or variable values within Figma.\n- [SAFE]: All included JavaScript helper scripts utilize standard Figma Plugin APIs and do not contain obfuscated code, unauthorized network calls, or suspicious dynamic execution patterns beyond their intended design system tasks.\n- [SAFE]: The skill does not attempt to escalate system privileges or establish persistence beyond the local temporary state file used for session recovery.
Audit Metadata