hookdeck-automation

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

• Third-party content exposure detected (high risk: 0.85). The required runtime workflow calls out to the Rube MCP server (RUBE_SEARCH_TOOLS, RUBE_MANAGE_CONNECTIONS, then RUBE_MULTI_EXECUTE_TOOL), and the LLM will ingest the returned tool schemas/status text from that external service, which is outsider-authored free text.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill requires configuring and calling the Rube MCP endpoint (https://rube.app/mcp) at runtime via RUBE_SEARCH_TOOLS/RUBE_MULTI_EXECUTE_TOOL so it fetches tool schemas and execution plans that directly control agent instructions and tool execution.