cloning-websites-to-weaverse

Pass

Audited by Gen Agent Trust Hub on Nov 13, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes the firecrawl tool to perform scraping and crawling operations on external websites provided by the user, saving results to the local .firecrawl/ directory.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data (HTML and markdown) retrieved from third-party websites. This data is used as a source for generating design specifications and implementation code.\n
  • Ingestion points: Data retrieved via firecrawl scrape and crawl operations, stored in .firecrawl/ JSON files.\n
  • Boundary markers: The instructions do not specify the use of delimiters or 'ignore' warnings for the ingested content.\n
  • Capability inventory: The agent is instructed to create and modify project files, including app/routes/clone-preview.$page.tsx, app/sections/ components, and app/weaverse/components.ts.\n
  • Sanitization: No explicit sanitization or content validation steps for the scraped data are defined.\n- [DYNAMIC_EXECUTION]: The workflow involves generating a functional React route at app/routes/clone-preview.$page.tsx based on the scraped content. This generated code is executed in the local development environment for the user to preview and approve the results before further development.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 13, 2026, 09:08 PM