The Agent Skills Directory

[DATA_EXPOSURE]: The skill provides guidance for managing the FIGMA_OAUTH_TOKEN environment variable, including instructions on how to persist it in shell profiles such as ~/.zshrc or ~/.bashrc. This is a routine procedure for configuring local development environment secrets.
[INDIRECT_PROMPT_INJECTION]: The skill processes design data and metadata retrieved from external Figma URLs which could theoretically contain malicious instructions.
Ingestion points: External design data is fetched via the get_design_context and get_metadata tools from Figma.
Boundary markers: The instructions do not specify the use of specific boundary markers or delimiters for the external data returned by the tools.
Capability inventory: The skill does not possess any dangerous capabilities such as arbitrary subprocess execution, file writing, or network operations beyond the defined Figma MCP tools.
Sanitization: There are no explicit sanitization or filtering steps mentioned for the content received from the Figma API before it is processed by the model.

figma