RATL.ai MCP

RATL.ai MCP provides intelligent, workflow-oriented tools for AI-powered test generation and execution. Access 25+ MCP tools for API testing and load testing through any MCP-compatible client.

What is RATL.ai MCP?

RATL.ai MCP (Model Context Protocol) server exposes comprehensive testing capabilities as MCP tools that AI assistants can use to:

• Generate API Tests: Automatically create pytest-based functional and E2E tests from your codebase
• Generate Load Tests: Automatically create Gatling-based performance tests
• Execute Tests: Run tests on cloud infrastructure and get detailed reports
• Analyze Results: Get AI-powered insights and recommendations

Available Tools

API Test Tools (10 tools):

• Codebase analysis for API testing
• Test plan generation (E2E scenarios + functional tests)
• Pytest code generation
• Test execution and reporting
• Local test result onboarding

Load Test Tools (15 tools):

• Codebase analysis for load testing
• Test plan generation (end-to-end scenarios)
• Gatling code generation
• Load test execution on cloud infrastructure
• Performance analysis and optimization
• Test comparison and benchmarking

Workflows

API Test Workflow:

Codebase Analysis → Test Plan → Code Generation → Validation → Execution → Reports

Load Test Workflow:

Codebase Analysis → Test Plan → Gatling Code → Validation → Execution → Analysis

For complete workflow documentation, see MCP.md.

MCP Server Endpoint

The RATL.ai MCP server is available at:

• Production: https://api.rattlez0.de/public/api/v1/mcp
• Development: http://localhost:8000/api/v1/mcp (if running locally)

Supported MCP Clients

RATL.ai MCP works with any MCP-compatible client:

• Cursor IDE - Native HTTP/SSE support (no bridge needed)
• Claude Desktop - Requires bridge (see setup below)
• Claude API - Native HTTP support
• OpenAI GPT Models - Native HTTP support
• Other MCP Clients - Use HTTP transport if supported, or bridge for stdio-based clients

Quick Start

Step 1: Get Your API Key

1. Visit ril.ratl.ai
2. Navigate to Settings > API Keys
3. Create a new API key or copy an existing one

Step 2: Configure Your MCP Client

Cursor IDE (Native HTTP Support)

1. Open Cursor Settings (Cmd/Ctrl + ,)
2. Navigate to Features > MCP
3. Click + Add New MCP Server
4. Configure:
   • Name: ratl.ai
   • Transport: Streamable HTTP (or SSE)
   • URL: https://api.rattlez0.de/public/api/v1/mcp
   • Headers: Add x-api-key header with your API key
5. Click Save

Claude Desktop (Requires Bridge)

Claude Desktop currently only supports stdio-based MCP servers. Use the bridge to connect:

Install the bridge:

# Option 1: Using npx (Recommended - No Installation)
# Just use npx in configuration (see below)

# Option 2: Install globally
npm install -g @ratl-ai/ratl-mcp

# Option 3: Direct download
# Download index.js from this repository

Configure Claude Desktop:

Location:

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json
• Linux: ~/.config/Claude/claude_desktop_config.json

Configuration (using npx - recommended):

{
  "mcpServers": {
    "ratl.ai": {
      "command": "npx",
      "args": [
        "-y",
        "@ratl-ai/ratl-mcp@latest",
        "https://api.rattlez0.de/public/api/v1/mcp",
        "--api-key",
        "YOUR_API_KEY_HERE"
      ]
    }
  }
}

Configuration (using global installation):

{
  "mcpServers": {
    "ratl.ai": {
      "command": "ratl-mcp",
      "args": [
        "https://api.rattlez0.de/public/api/v1/mcp",
        "--api-key",
        "YOUR_API_KEY_HERE"
      ]
    }
  }
}

Configuration (using direct path):

{
  "mcpServers": {
    "ratl.ai": {
      "command": "node",
      "args": [
        "/path/to/mcp-bridge/index.js",
        "https://api.rattlez0.de/public/api/v1/mcp",
        "--api-key",
        "YOUR_API_KEY_HERE"
      ]
    }
  }
}

Restart Claude Desktop after configuration.

Other MCP Clients

For clients that support HTTP-based MCP:

{
  "mcpServers": {
    "ratl.ai": {
      "url": "https://api.rattlez0.de/public/api/v1/mcp",
      "transport": "sse",
      "headers": {
        "x-api-key": "YOUR_API_KEY_HERE"
      }
    }
  }
}

For clients that only support stdio, use the bridge tool (see Claude Desktop setup above).

Example Use Cases

Creating API Tests

You: "Help me create API tests for the OTT API at https://ott.ratl.app"

AI Assistant (using RATL.ai MCP tools):

1. Analyzes your codebase/OpenAPI spec
2. Generates E2E test scenarios and functional test plan
3. Creates pytest test files
4. Validates the code
5. Executes tests and provides reports

Creating Load Tests

You: "Generate load tests for my authentication flow"

AI Assistant (using RATL.ai MCP tools):

1. Analyzes your API endpoints
2. Creates a comprehensive end-to-end load test scenario
3. Generates Gatling Java code
4. Validates code and URL
5. Executes load test on cloud infrastructure
6. Provides performance analysis and recommendations

What You Get

API Testing

• E2E Test Scenarios: User journey tests that test complete workflows
• Functional Tests: Individual endpoint tests with comprehensive coverage
• Automatic Code Generation: AI generates pytest code from your API
• Validation: Automatic code validation before execution
• Detailed Reports: E2E and functional test reports with request/response details

Load Testing

• End-to-End Scenarios: Complete user flows tested under load
• Gatling Code Generation: Production-ready Gatling Java code
• Cloud Execution: Tests run on scalable cloud infrastructure
• Performance Metrics: Response times, throughput, error rates
• AI Analysis: Automated insights and optimization recommendations

Authentication

Authentication is handled via API key:

1. Get API Key: From ril.ratl.ai Settings > API Keys
2. Pass in Configuration:
   • For HTTP clients: Include x-api-key header
   • For bridge: Include --api-key YOUR_API_KEY in args
3. Session Tokens: The server supports session-based authentication (tokens valid for 1 hour)

Session-Based Authentication (Recommended)

Step 1: Initialize with API key

POST /api/v1/mcp
Headers: x-api-key: your-api-key
Body: {
  "method": "initialize",
  "params": {"api_key": "your-api-key"}
}
Response: {
  "result": {
    "sessionToken": "abc123...",
    "authenticated": true
  }
}

Step 2: Use session token for subsequent calls

POST /api/v1/mcp
Headers: x-session-token: abc123...
Body: {
  "method": "tools/call",
  "params": {...}
}

Troubleshooting

"command not found: npx" or "command not found: node"

Solution:

• Install Node.js: https://nodejs.org/
• Or use direct path to index.js in configuration

"HTTP 401: Unauthorized"

Solution:

• Verify your API key is correct
• Check API key hasn't expired
• Get a new API key from ril.ratl.ai

Bridge Not Working (Claude Desktop)

Solution:

1. Test bridge manually:

   echo '{"jsonrpc":"2.0","id":1,"method":"tools/list","params":{}}' | \
     npx @ratl-ai/ratl-mcp@latest https://api.rattlez0.de/public/api/v1/mcp --api-key YOUR_KEY

2. Check client logs for errors
3. Verify file paths are absolute (not relative)

Tools Not Appearing

Solution:

1. Restart your MCP client after configuration changes
2. Verify API key is correct
3. Check client logs for connection errors
4. For HTTP clients, verify the URL is accessible

Bridge Tool (For stdio-based MCP Clients)

The bridge is a technical implementation that enables stdio-based MCP clients (like Claude Desktop) to connect to RATL.ai's HTTP-based MCP server:

1. Client sends JSON-RPC requests via stdin
2. Bridge forwards requests to RATL.ai MCP server over HTTP
3. Bridge receives responses and writes to stdout
4. Client reads responses and processes them

Note: When your MCP client adds native HTTP MCP support, the bridge will no longer be needed.

Bridge Installation

# Install globally
npm install -g @ratl-ai/ratl-mcp

# Or use npx (no installation)
npx @ratl-ai/ratl-mcp@latest <url> [options]

Bridge Command Line Options

ratl-mcp <url> [options]

Options:
  --api-key <key>        API key for authentication (sets x-api-key header)
  --header <name:value>  Additional header (can be used multiple times)
  --transport <type>     Transport type (default: http)

Documentation

• Complete MCP Documentation: See MCP.md for detailed tool documentation, workflows, and examples
• API Documentation: See README.md for REST API documentation
• Website: https://ratl.ai

Support

• Website: https://ratl.ai
• Get API Keys: https://ril.ratl.ai (Settings > API Keys)
• Documentation: MCP.md - Complete MCP server documentation

License

MIT License - See LICENSE file for details.

RATL.ai - Intelligent testing tools powered by AI. Generate, execute, and analyze tests with ease.