| FEATURE | DESCRIPTION |
|---|---|
Native agent interoperability via MCP |
Cortex Agentic Assistant now supports native interaction with external environments via the Model Context Protocol (MCP). Cortex AI Agents can seamlessly communicate with third-party MCP servers to discover available tools and automatically generate agentic actions, enabling you to automate more complex security tasks and accelerate incident response. |
Increased transparency into agent execution |
Gain visibility into your AI workflows by accessing the specific data generated when agents create objects or retrieve information. JSON artifacts are available directly in the agent’s plan view to provide technical context for results. |
Manage your AI prompts in one place with the Prompt Library |
Speed up your development process and ensure consistency by using a single library to manage all your AI prompts. You can now add these prompts directly to playbooks as AI tasks or register them as specific agent actions. To keep your workspace clean, the Prompt Library remains hidden unless Agentic Assistant is enabled in the server settings. |
Cortex Agentic Assistant new supported regions |
The Cortex Agentic Assistant is currently available for tenants in the following regions:
|
Automation Engineer Agent |
Accelerate script creation through an intuitive, conversational chat experience. Use natural language prompts to build and refine secure, high-quality Python scripts. The agent applies security best practices and uses the Agentix Automation SDK to ensure high-quality results. |