semanticscholar-automation

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill requires connecting to the Rube MCP endpoint (https://rube.app/mcp) at runtime and calls RUBE_SEARCH_TOOLS to fetch tool schemas, slugs, and recommended execution plans which directly control what instructions/tools the agent uses and thus can execute remote-controlled behavior.