AI Agent Integration

TestKase MCP Server

Connect AI agents like Claude, GitHub Copilot, and Cursor to TestKase via the Model Context Protocol. Create test cases, run cycles, and pull reports — all through natural language.

@testkase/mcp-serveron npm

Get started free View documentation

Model Context Protocol

What is MCP?

The Model Context Protocol (MCP) is an open standard that lets AI agents interact with external tools and services. Instead of copying and pasting between your AI agent and TestKase, the MCP server gives your agent direct access to create test cases, run test cycles, pull reports, and more — all through natural conversation.

Install

Add @testkase/mcp-server to your AI agent config — runs via npx, no global install needed.

Authenticate

Set your TestKase PAT token as an environment variable.

Talk

Ask your agent to manage tests in plain English.

Supported AI Agents

Works out of the box with the most popular MCP-compatible AI agents.

Claude Desktop

Anthropic's desktop app for Claude — full MCP support built in.

GitHub Copilot

Agent mode in VS Code — use MCP tools directly in your editor.

Claude Code

Anthropic's CLI agent — add the MCP server with a single command.

Cursor

AI-powered code editor with native MCP support via project config.

11 Built-in Tools

What Your Agent Can Do

Full test lifecycle management through natural language — no API calls to memorize.

Create Test Cases

Generate test cases with steps, priorities, labels, and folder placement — from a single prompt.

Search & Filter

Find test cases by keyword, status, priority, label, or folder — with sorting and pagination.

Run Test Cycles

Create cycles, link test cases, assign testers, and manage the full cycle lifecycle.

Execute Tests

Record pass, fail, blocked, or skipped results — single or bulk execution.

Pull Reports

Access 40+ report types: execution summaries, coverage, trends, defects, and AI insights.

Manage Structure

Create, rename, move, or delete folders. Manage test plans and link cycles to plans.

Setup Guide

Choose your AI agent and follow the setup instructions. Takes less than a minute.

Add to your config file at %APPDATA%\Claude\claude_desktop_config.json (Windows) or ~/Library/Application Support/Claude/claude_desktop_config.json (macOS):

{
  "mcpServers": {
    "testkase": {
      "command": "npx",
      "args": ["-y", "@testkase/mcp-server"],
      "env": {
        "TESTKASE_API_BASE_URL": "https://api.testkase.com",
        "TESTKASE_PAT_TOKEN": "YOUR_PAT_TOKEN"
      }
    }
  }
}

Replace YOUR_PAT_TOKEN with your actual PAT. Restart Claude Desktop after saving.

Example Conversations

Just tell your AI agent what you need in plain English.

"Create 5 test cases for the checkout flow covering payment, shipping, promo codes, cart updates, and order confirmation"

Bulk creates five structured test cases with steps, each organized under the appropriate folder.

"Create a folder called Regression under the root and move all high-priority test cases into it"

Creates a new folder, searches for high-priority test cases, and moves them into the Regression folder.

"Create a new test cycle called Sprint 42 QA and add all test cases from the Login folder"

Creates the cycle, searches for test cases in the Login folder, and links them to the new cycle.

"Execute TEST-5, TEST-6, and TEST-7 as passed in TCYCLE-10"

Bulk-executes three test cases with a pass result in the specified cycle.

"Show me a risk heatmap by folder for PRJ-1001"

Pulls the risk_heatmap_folder report showing risk distribution across your folder structure.

"Create a test plan for the v2.0 release and link the Sprint 41 and Sprint 42 cycles to it"

Creates a test plan, then links two existing test cycles to it for release tracking.

"Find all blocked test cases in the Payments folder"

Searches test cases filtered by execution status (blocked) within the Payments folder.

Start Testing with AI

Get Started Now