figma-design-to-code
Pass
Audited by Gen Agent Trust Hub on Oct 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external design data retrieved from Figma via the TemPad Dev MCP tool, which serves as a potential vector for indirect instructions.
- Ingestion points: Design data is fetched using
get_codeandget_structuretool calls, as described inSKILL.mdandreferences/recovery.md. - Boundary markers: The instructions lack explicit boundary markers or warnings to the agent to disregard text content within the design that may resemble commands.
- Capability inventory: The skill allows for modification of local project files and the execution of project-defined scripts and checks as outlined in Step 5 of
SKILL.md. - Sanitization: No sanitization or filtering of the incoming design data is specified before it is processed by the agent.
- [COMMAND_EXECUTION]: The skill instructions in Step 5 direct the agent to execute local project scripts and checks for verification purposes. While this is a standard feature for development tools, it involves the execution of commands defined within the local project environment.
Audit Metadata