Swimlane Adds Custom Detections and Two Way MCP Support to Turbine

Swimlane has added new Turbine capabilities for custom security detections, two way Model Context Protocol support and application building through natural language prompts, according to a release published by Cybersecurity Insiders. The October 2026 update is generally available.

Security teams can create detection logic for activity such as prompt injection, unsanctioned AI accounts, sensitive data submitted to AI services and disabled security guardrails. Swimlane says a single big data action can evaluate tens of thousands of records across multiple compute nodes, and Turbine can route resulting alerts to Swimlane AI SOC for investigation and response.

Turbine can operate as an MCP host, allowing Hero AI to access tools from external MCP servers. It can also act as an MCP server, letting assistants including Claude, ChatGPT, Cursor, Gemini CLI and Microsoft Copilot invoke Turbine playbooks under role based access controls.

The Hero AI Application Builder Agent can generate or modify applications, record structures, layouts and workflows from user instructions. Proposed changes are highlighted for human review and are not saved until a user accepts them.