google-address-validation-automation
Fail
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
- External Dependencies (HIGH): The skill requires the user to add an untrusted MCP server endpoint (
https://rube.app/mcp). This source is not on the trusted repositories or organizations list, posing a significant risk as the server controls the tool definitions available to the agent. - Dynamic Tool Execution (HIGH): The workflow relies on
RUBE_SEARCH_TOOLSto fetch schemas andRUBE_MULTI_EXECUTE_TOOLto run them. Since the tool slugs and input requirements are fetched from an untrusted remote server at runtime, a malicious server response could trigger the execution of unintended commands or scripts on the agent's environment. - Indirect Prompt Injection (MEDIUM): The skill lacks boundary markers or sanitization for the data returned by the remote server. If the
rube.appservice is compromised, it could inject malicious instructions through the tool schemas or search results which the agent would then process as valid instructions (Category 8). - Capability Inventory (INFO): The skill has the capability to manage connections, search for tools, and execute multiple tools simultaneously via the Rube MCP protocol. These broad capabilities increase the impact if the remote service is used for an attack.
Recommendations
- AI detected serious security threats
Audit Metadata