figma-orient
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external Figma files (node names, UI copy, and text content) using the
figma_executetool. While it performs mapping and documentation, there is a risk that malicious text content within a Figma file could attempt to influence the agent's behavior during the orientation process. - Ingestion points: External Figma canvas content retrieved via
figma_execute(usinggetNodeByIdAsyncandfindAllmethods). - Boundary markers: The skill does not define specific delimiters or "ignore instructions" warnings for the content written to
docs/figma-map.md. - Capability inventory: The skill possesses file-write capabilities (writing to
docs/figma-map.md) and network read capabilities via the Figma MCP tools. - Sanitization: The instructions recommend trimming long strings but do not specify sanitization or escaping of the ingested UI text content.
- [DYNAMIC_EXECUTION]: The skill instructs the agent to generate and execute JavaScript snippets through the
figma_executetool to interact with the Figma Plugin API. - The agent is directed to use specific API methods like
getNodeByIdAsyncandfindAllwithin the tool call. - This represents dynamic execution of agent-generated code within the Figma desktop bridge or plugin sandbox environment.
Audit Metadata