reply-io-automation
Warn
Audited by Snyk on Feb 16, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).
- Potentially malicious external URL detected (high risk: 0.90). The skill explicitly requires adding and using the MCP server at https://rube.app/mcp at runtime, and RUBE_SEARCH_TOOLS / RUBE_MANAGE_CONNECTIONS calls will fetch tool schemas and execution plans from that endpoint which directly control what prompts/instructions and remote tool code the agent runs.
Audit Metadata