langbase-automation

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.80). The skill explicitly requires calling RUBE_SEARCH_TOOLS against the public Rube MCP endpoint (https://rube.app/mcp), which returns third-party tool schemas and recommended execution plans that the agent is expected to read and use to decide and execute actions, allowing untrusted content to influence behavior.

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 MCP server endpoint https://rube.app/mcp at runtime to discover tool schemas and receive recommended execution plans and to execute RUBE_MULTI_EXECUTE_TOOL, so this external URL is a required runtime dependency that can control agent behavior and trigger remote code/tool execution.