Zephyr Enterprise MCP

The package runs as a Model Context Protocol (MCP) server, allowing AI assistants like Claude, GitHub Copilot, and Cursor to query Zephyr Enterprise directly — no manual data export or copy-pasting required. Configuration takes two environment variables (ZEPHYR_BASE_URL and ZEPHYR_TOKEN); project and release IDs are passed per call, so one server instance works across any number of projects.

All 15 tools are available once the server is connected:

Projects & Releases

• list_projects — List all accessible Zephyr projects
• list_releases — List all releases for a given project
• project_health — Overall health score (0–100) with metrics and recommendations

Test Cycles

• list_cycles — All test cycles for a release, including phases and execution status counts
• get_cycle — Full detail for a single test cycle

Test Cases & Coverage

• search_test_cases — Search test cases by keyword query
• test_coverage — Detailed test coverage analysis
• req_coverage — Requirements with and without test coverage
• failed_tests — List and analyze failed test cases

Trends & Charts

• test_trends — Execution trends over time with daily breakdown
• execution_burnup — Day-by-day cumulative executed vs ideal and total scope
• execution_burndown — Day-by-day remaining vs ideal test counts

Users & Activity

• user_activity — Team productivity metrics — who is assigned and executing tests
• user_trend — Full audit log for a specific user, filterable by date range, entity, and operation (requires full email address)
• list_users — All users assigned to a project with name, email, role, and status

MCP Integration

Run Zephyr Enterprise Tools as a Model Context Protocol (MCP) server to give an AI assistant direct access to Zephyr Enterprise data. Once connected, users can ask questions such as: “Show me all failing tests in release 4312.”
The assistant can then invoke the appropriate Zephyr tool to retrieve the requested information.

Claude Desktop
Edit your Claude Desktop configuration file, claude_desktop_config.json, and add the Zephyr server:

{
  "mcpServers": {
    "zephyr": {
      "command": "npx",
      "args": [
        "-y",
        "--package",
        "zephyr-enterprise-tools@latest",
        "zephyr-mcp"
      ],
      "env": {
        "ZEPHYR_BASE_URL": "https://your-zephyr.com/flex/services/rest/latest",
        "ZEPHYR_TOKEN": "your-api-token"
      }
    }
  }
}

VS Code with GitHub Copilot
Create or edit .vscode/mcp.json in the workspace

{
  "servers": {
    "zephyr": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "--package",
        "zephyr-enterprise-tools",
        "zephyr-mcp"
      ],
      "env": {
        "ZEPHYR_BASE_URL": "https://your-zephyr.com/flex/services/rest/latest",
        "ZEPHYR_TOKEN": "your-api-token"
      }
    }
  }
}

Cursor
Add the following server configuration to mcp.json:

{
  "mcpServers": {
    "zephyr-enterprise": {
      "command": "npx",
      "args": ["-y", "zephyr-enterprise-tools@latest", "mcp"],
      "env": {
        "ZEPHYR_BASE_URL": "https://your-zephyr.com/flex/services/rest/latest",
        "ZEPHYR_TOKEN": "your-api-token"
      }
    }
  }
}