figma-click-flow
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DYNAMIC_EXECUTION]: The skill uses a template-based code generation approach, substituting parameters into a JavaScript snippet (
references/draw-overlay.js) which is then executed via thefigma_executetool. This method is used to perform complex operations in the Figma document. \n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the Figma canvas, including node names and prototype reactions, which are then used in the construction of new layer names and validation screenshots. This creates a potential surface where malicious content in a Figma file could attempt to influence the agent's behavior. \n - Ingestion points: Node names and reaction metadata retrieved from the Figma section descendants. \n
- Boundary markers: None; the skill does not use delimiters or instructions to ignore embedded content in processed node data. \n
- Capability inventory: The skill utilizes
figma_executefor document modification andfigma_capture_screenshotfor state verification. \n - Sanitization: No sanitization or escaping is performed on external node data before it is incorporated into layer names or SVG structures.
Audit Metadata