design-context-extract
Pass
Audited by Gen Agent Trust Hub on May 14, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external URLs and screenshots during the design extraction process.
- Ingestion points: Uses
WebFetchfor live URLs and theReadtool for image files (multimodal analysis). - Boundary markers: No specific boundary markers or instructions to ignore embedded textual content are provided to the agent.
- Capability inventory: The skill has access to
Bash,Write, andWebFetchtools. - Sanitization: No explicit sanitization or validation of extracted content beyond the internal model logic is mentioned.
- [COMMAND_EXECUTION]: The instructions suggest recommending specific CLI commands to the user, such as
npx shadcn@latest apply luma. These commands are related to well-known, industry-standard UI tools and are presented as user recommendations rather than automated background execution.
Audit Metadata