design-extract
Pass
Audited by Gen Agent Trust Hub on Aug 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from multiple sources to extract design tokens.
- Ingestion points: Processes source code (
CSS,Tailwind config,styled-components), Figma export files (figma/tree.json), and screenshot images. - Boundary markers: The skill lacks explicit boundary markers or instructions for the agent to ignore natural language instructions that might be embedded within code comments or design metadata.
- Capability inventory: The skill possesses the
fs:readcapability to access the local filesystem andprompt:injectto modify the agent's context. - Sanitization: No specific sanitization or validation logic is defined for the ingested files before they are interpreted by the agent.
Audit Metadata