clone-website
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill extracts content from external URLs via browser tools. Ingestion occurs through browser MCP tools reading DOM and text. There are no specified boundary markers or sanitization processes, creating a vulnerability surface for instructions embedded in target sites. Capabilities include file writes, command execution, and script running.
- [DYNAMIC_EXECUTION]: The skill executes dynamic JS snippets via browser MCP tools for styling extraction and discovers assets. It also generates and runs custom .mjs scripts for asset downloading.
- [COMMAND_EXECUTION]: The skill runs npm run build and npx tsc --noEmit as verification steps.
- [EXTERNAL_DOWNLOADS]: The skill downloads images, videos, and font assets from target URLs to the local public directory.
Audit Metadata