build-design
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources (Figma files and design images) to generate project code and save assets to the filesystem.
- Ingestion points: External Figma URLs and design files are read using MCP tools like
get_design_contextandget_metadatainSKILL.mdandfigma.md. - Boundary markers: The instructions do not specify any boundary markers or delimiters to separate the design data from the agent's instructions, nor do they instruct the agent to ignore instructions embedded within the design content (e.g., layer names, text fields, or variable values).
- Capability inventory: The agent has the capability to write files to the local project (
Save each asset into the project) and modify the codebase by implementing UI components. - Sanitization: There is no mention of sanitizing or validating the data retrieved from Figma before it is used to generate code or write files.
Audit Metadata