clone-website

Warn

Audited by Gen Agent Trust Hub on Apr 24, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes several shell commands to build and verify the cloned project, including npm run build, npx tsc --noEmit, and the execution of a dynamically created Node.js asset download script.
  • [EXTERNAL_DOWNLOADS]: It identifies and downloads remote assets such as images, videos, and fonts from the target URL provided by the user to the local filesystem.
  • [REMOTE_CODE_EXECUTION]: The skill scrapes logic, structure, and content from a remote website and uses it to generate local source code and scripts which are then executed or compiled within the user's environment.
  • [PROMPT_INJECTION]: The skill processes untrusted web data via browser automation to drive the generation of code and specifications, presenting a surface for indirect prompt injection.
  • Ingestion points: Data enters the system via browser MCP tool access to the user-supplied target URL.
  • Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions for the extracted content when passed to builder agents.
  • Capability inventory: The skill has file system access, the ability to write code, and the capability to execute shell commands.
  • Sanitization: There is no documented validation or sanitization process for the text and CSS content extracted from the target website before it is used in code generation.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 24, 2026, 05:43 PM
Security Audit — agent-trust-hub — clone-website