wati-automation

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTIONREMOTE_CODE_EXECUTION
Full Analysis
  • EXTERNAL_DOWNLOADS (LOW): The skill requires connection to an external MCP server at https://rube.app/mcp. This domain is not part of the trusted organization list, making it an unverified remote dependency.\n- PROMPT_INJECTION (LOW): Category 8: Indirect Prompt Injection surface. \n
  • Ingestion points: The agent fetches tool schemas and execution plans dynamically from RUBE_SEARCH_TOOLS.\n
  • Boundary markers: None; instructions direct the agent to follow the retrieved schemas and execution plans strictly.\n
  • Capability inventory: The skill includes tools for multi-tool execution and a remote workbench (RUBE_REMOTE_WORKBENCH).\n
  • Sanitization: No input validation or sanitization of the remote schema data is mentioned.\n- REMOTE_CODE_EXECUTION (LOW): The skill utilizes RUBE_REMOTE_WORKBENCH to perform operations that are defined by the remote server's responses.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:36 PM