Vibe Coded with Claude Code and Pu-Er 🍵
A Model Context Protocol (MCP) server that provides LLM access to Cucumber Studio's testing platform. This server enables AI assistants to retrieve test scenarios, action words, test runs, and project information from Cucumber Studio.
- Dual Transport Support - STDIO and Streamable HTTP transports with session management
- Project Management - List and retrieve project details
- Scenario Access - Browse test scenarios and search by tags
- Action Words - Access reusable test steps and definitions
- Test Execution - View test runs, executions, and build information
- Hot Reload Development - Instant server restart on file changes with tsx --watch
- Configurable Logging - Structured logging with multiple output destinations
- Comprehensive Error Handling - Robust error handling with detailed feedback
- Type Safety - Full TypeScript implementation with Zod validation
- Comprehensive Testing - 82%+ test coverage with Vitest and MSW
The easiest way to use this MCP server is as a Desktop Extension:
-
Download Extension: Get the latest
.dxt
file from the releases page (automatically built from each release) - Install Extension: Import the extension in your compatible AI desktop application
-
Configure Credentials: Set up your Cucumber Studio API credentials through the extension settings:
- Access Token: Your Cucumber Studio API access token
- Client ID: Your Cucumber Studio client ID
- User ID: Your Cucumber Studio user ID
The extension will automatically handle the MCP server setup and communication.
Run directly with npx (no installation required):
npx cucumberstudio-mcp
Set your environment variables first:
export CUCUMBERSTUDIO_ACCESS_TOKEN="your_token"
export CUCUMBERSTUDIO_CLIENT_ID="your_client_id"
export CUCUMBERSTUDIO_UID="your_uid"
- Clone the repository:
git clone https://github.com/HeroSizy/cucumberstudio-mcp.git
cd cucumberstudio-mcp
- Install dependencies:
npm install
- Set up environment variables:
cp .env.example .env
# Edit .env with your Cucumber Studio API credentials
- Build the server:
npm run build
Run the official Docker image from Docker Hub:
# With environment file
docker run --env-file .env herosizy/cucumberstudio-mcp
# With environment variables
docker run -e CUCUMBERSTUDIO_ACCESS_TOKEN=your_token \
-e CUCUMBERSTUDIO_CLIENT_ID=your_client_id \
-e CUCUMBERSTUDIO_UID=your_uid \
herosizy/cucumberstudio-mcp
- Set up environment variables:
cp .env.example .env
# Edit .env with your Cucumber Studio API credentials
- Update docker-compose.yml to use the pre-built image:
version: '3.8'
services:
cucumberstudio-mcp:
image: herosizy/cucumberstudio-mcp
env_file:
- .env
restart: unless-stopped
ports:
- "${MCP_PORT:-3000}:3000"
- Run with Docker Compose:
docker-compose up
- Build the image:
npm run docker:build
- Run the container:
npm run docker:run
The Docker setup includes health checks and automatic restarts for production use. The multi-stage build process creates optimized production images with only runtime dependencies (~150MB).
The server requires Cucumber Studio API credentials. Get these from your Cucumber Studio account settings:
-
CUCUMBERSTUDIO_ACCESS_TOKEN
- Your API access token -
CUCUMBERSTUDIO_CLIENT_ID
- Your client ID -
CUCUMBERSTUDIO_UID
- Your user ID
-
CUCUMBERSTUDIO_BASE_URL
- API base URL (default: https://studio.cucumberstudio.com/api) -
MCP_TRANSPORT
- Transport type:stdio
(default),http
, orstreamable-http
-
MCP_PORT
- HTTP transport port (default: 3000) -
MCP_HOST
- HTTP transport host (default: 0.0.0.0) -
MCP_CORS_ORIGIN
- CORS origin setting (default: true)
-
LOG_LEVEL
- Log level:error
,warn
,info
,debug
,trace
(default: info) -
LOG_API_RESPONSES
- Log Cucumber Studio API responses (default: false) -
LOG_REQUEST_BODIES
- Log API request bodies for debugging (default: false) -
LOG_RESPONSE_BODIES
- Log API response bodies for debugging (default: false) -
LOG_TRANSPORT
- Logging output:console
,stderr
,file
,none
(default: stderr) -
LOG_FILE
- Log file path (required if LOG_TRANSPORT=file)
The server supports both STDIO and HTTP transports:
# Development
npm run dev
# Production
npm start
# Development
npm run dev:http
# Production
npm run start:http
Import the .dxt
extension file directly into your compatible AI desktop application. The extension handles all configuration through its settings interface.
For manual MCP client configuration:
{
"mcpServers": {
"cucumberstudio": {
"command": "npx",
"args": ["cucumberstudio-mcp"],
"env": {
"CUCUMBERSTUDIO_ACCESS_TOKEN": "your_token",
"CUCUMBERSTUDIO_CLIENT_ID": "your_client_id",
"CUCUMBERSTUDIO_UID": "your_uid"
}
}
}
}
{
"mcpServers": {
"cucumberstudio": {
"command": "node",
"args": ["/path/to/cucumberstudio-mcp/build/index.js"],
"env": {
"CUCUMBERSTUDIO_ACCESS_TOKEN": "your_token",
"CUCUMBERSTUDIO_CLIENT_ID": "your_client_id",
"CUCUMBERSTUDIO_UID": "your_uid"
}
}
}
}
{
"mcpServers": {
"cucumberstudio": {
"command": "docker",
"args": ["run", "--rm", "-i", "--env-file", "/path/to/.env", "herosizy/cucumberstudio-mcp"]
}
}
}
{
"mcpServers": {
"cucumberstudio": {
"command": "docker",
"args": ["run", "--rm", "-i", "--env-file", "/path/to/.env", "cucumberstudio-mcp"]
}
}
}
-
cucumberstudio_list_projects
- List all accessible projects -
cucumberstudio_get_project
- Get detailed project information
-
cucumberstudio_list_scenarios
- List scenarios in a project -
cucumberstudio_get_scenario
- Get detailed scenario information -
cucumberstudio_find_scenarios_by_tags
- Find scenarios by tags
-
cucumberstudio_list_action_words
- List reusable action words -
cucumberstudio_get_action_word
- Get detailed action word information -
cucumberstudio_find_action_words_by_tags
- Find action words by tags
-
cucumberstudio_list_test_runs
- List test runs -
cucumberstudio_get_test_run
- Get detailed test run information -
cucumberstudio_get_test_executions
- Get individual test results -
cucumberstudio_list_builds
- List builds -
cucumberstudio_get_build
- Get build details -
cucumberstudio_list_execution_environments
- List execution environments
The server supports hot reload for rapid development:
# STDIO transport with hot reload
npm run dev
# HTTP transport with hot reload
npm run dev:http
Files are automatically recompiled and the server restarts when changes are detected.
# Install dependencies
npm install
# Run type checking
npm run typecheck
# Run linting
npm run lint
# Build for production
npm run build
# Run tests
npm test
# Run tests in watch mode
npm run test:watch
# Run tests with coverage (82%+ coverage)
npm run test:coverage
# Run tests with UI
npm run test:ui
# Production build (default) - optimized for size, no .d.ts/.js.map files
npm run build
# Development build - includes source maps and type declarations for debugging
npm run build:dev
# Validate manifest.json
npm run dxt:validate
# Build complete DXT extension for local testing (optimized production build)
npm run dxt:build
# Check info about built extension
npm run dxt:info
# Clean up build artifacts
npm run dxt:clean
The server is built with a modular, production-ready architecture:
- TypeScript - Full type safety with strict configuration
- Dual Transports - STDIO for local use, Streamable HTTP for remote access
- Zod - Runtime validation for API inputs and configuration
- Axios - HTTP client with comprehensive error handling and logging
- MCP SDK - Official Model Context Protocol implementation
- Express - HTTP server with CORS, security middleware, and session management
- Vitest - Modern testing framework with 82%+ code coverage
- MSW - Mock Service Worker for realistic API testing
- Session Management - HTTP transport with session tracking and cleanup
- Comprehensive Logging - Structured logging with configurable outputs and levels
- Error Handling - Robust error handling with detailed feedback and recovery
- Security - Origin validation, CORS protection, and input sanitization
- Health Monitoring - Health check endpoints and request/response tracking
- Development Workflow - Hot reload, comprehensive testing, and Docker support
The project includes comprehensive test coverage:
# Run all tests
npm test
# Run tests with coverage report
npm run test:coverage
# Run tests in watch mode (for development)
npm run test:watch
Test coverage includes:
- Unit tests for all modules
- Integration tests for the MCP server
- Transport layer testing
- API client mocking and testing
- Configuration validation
- Error handling scenarios
This project uses automated releases via GitHub Actions. When a version tag is pushed, it automatically:
- Runs full test suite - Ensures code quality and coverage
-
Publishes to NPM - Makes the package available via
npx cucumberstudio-mcp
- Builds and publishes Docker image - Pushes multi-platform images to Docker Hub
- Creates GitHub release - Generates release notes and links
- Update the version in
package.json
:
npm version patch|minor|major
- Push the tag to trigger the release:
git push origin --tags
- The GitHub Action will automatically:
- Publish to NPM: https://www.npmjs.com/package/cucumberstudio-mcp
- Push to Docker Hub: https://hub.docker.com/r/herosizy/cucumberstudio-mcp
- Create a GitHub release with changelog
For automated publishing, the following secrets must be configured in the GitHub repository:
-
NPM_TOKEN
- NPM authentication token -
DOCKER_USERNAME
- Docker Hub username -
DOCKER_PASSWORD
- Docker Hub password or access token
- Fork the repository
- Create a feature branch
- Make your changes
- Add tests for new functionality
- Ensure all tests pass:
npm test
- Submit a pull request
MIT License - see LICENSE file for details