Overview

The Zowe CLI MCP is a Model Context Protocol server that bridges AI assistants with z/OS job management via the Zowe SDK. It enables conversational agents—such as Claude—to query real‑time job information from z/OS MF profiles without requiring direct access to the mainframe. By exposing a lightweight, standardized interface, this MCP lets developers embed z/OS job insights into chat‑based workflows, monitoring dashboards, or automated support systems.

Problem Solved

Mainframe operators and developers often need to inspect job queues, track user activity, or troubleshoot batch failures. Traditional methods rely on command‑line utilities or proprietary interfaces that are not easily consumable by AI assistants. The Zowe CLI MCP solves this gap by providing a declarative API that translates common job‑query operations into structured responses. This eliminates the need for custom scripts or manual parsing, reducing cognitive load and speeding up issue resolution.

Server Functionality

The server offers two primary tools:

• get-user-jobs – Returns the full list of jobs submitted by a specified user on a given z/OS profile.
• get-jobs-by-name – Retrieves jobs whose names match a provided prefix, again scoped to a particular profile.

Both tools leverage the Zowe SDK’s robust job‑management capabilities, ensuring that queries are accurate and up‑to‑date. The server handles authentication, profile resolution, and response formatting automatically, allowing the AI client to focus on intent extraction and result presentation.

Key Features & Capabilities

• Zowe SDK Integration – Direct use of Zowe’s proven APIs guarantees compatibility with modern z/OS MF environments.
• Profile Awareness – Users can target specific mainframe configurations without hard‑coding connection details.
• Structured Output – Results are returned as JSON objects, making them easy to consume, filter, or display in UI components.
• Extensibility – The MCP framework allows future tools (e.g., job status, history) to be added with minimal friction.

Use Cases

• AI‑powered DevOps: A chatbot can answer “Show me all jobs for user ” and immediately display the list, aiding rapid triage.
• Automated Monitoring: Integrate the MCP into a continuous‑integration pipeline that checks for orphaned jobs before deployment.
• Support Ticketing: Customer service agents can ask the AI for job details related to a reported issue, reducing back‑and‑forth with mainframe specialists.

Integration with AI Workflows

Because the MCP adheres to the standard Model Context Protocol, any Claude‑compatible client can invoke these tools simply by including a structured request in the conversation. The AI processes the user’s natural language, maps it to the appropriate tool, and streams back a concise summary or full job list. This seamless interaction turns complex mainframe queries into conversational commands, dramatically improving developer productivity and reducing the learning curve for non‑mainframe users.

By encapsulating z/OS job queries behind a clean, AI‑friendly interface, the Zowe CLI MCP empowers developers and operators to harness mainframe data within modern conversational platforms—making legacy systems more accessible, responsive, and integrated into today’s automated workflows.