The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill facilitates a workflow where the agent ingests and processes data from external Figma designs via the Chrome MCP and Figma MCP tools, which can be manipulated by an attacker to include malicious instructions.
Ingestion points: The 'Figma Extraction Pipeline' utilizes Chrome MCP to navigate to Figma URLs and Figma MCP to extract colors, typography, and layout elements.
Boundary markers: The instructions do not define clear delimiters or instruct the agent to ignore potentially malicious natural language instructions embedded within the Figma design's text layers or metadata.
Capability inventory: The skill instructs the agent to generate SCSS and Python configuration code based on the extracted data and utilizes the WebFetch tool for browser-based extraction.
Sanitization: No sanitization, validation, or filtering of the content extracted from external Figma designs is described before it is used in code generation or prompt construction.

theme-design