How to properly configure and deploy the damn-vulnerable-MCP-server for local penetration testing? I’ve followed the README steps, but the server fails to start with a ‘Port 8080 already in use’ error. Are there specific dependency requirements or recommended isolation practices (e.g., Docker) to avoid conflicts? Additionally, could you clarify the intended vulnerabilities in the MCP protocol implementation for focused testing (e.g., buffer overflow, auth flaws)?