Test Genie

Test Genie MCP is an AI-powered MCP server that automates the full app testing pipeline, covering scenario generation, test execution, issue detection, fix suggestions, and automated applications. It supports multiple platforms and integrates AI-assisted workflows to help you efficiently validate iOS, Android, Flutter, React Native, and Web applications.

How to use

You interact with Test Genie MCP through a client that can issue the platform-agnostic MCP commands described in your workflow. You can start by analyzing your app, generating test scenarios, running tests, detecting issues, and applying AI-suggested fixes. The goal is to create an end-to-end automated testing loop that reduces manual effort while increasing coverage and reliability.

How to install

Prerequisites you should have before installing Test Genie MCP:

• Ensure Node.js is installed (node -v) and npm is available (npm -v)

Install Test Genie MCP globally or run from source depending on your preference.

npm install -g test-genie-mcp

If you prefer to build from source, follow these steps:

git clone https://github.com/MUSE-CODE-SPACE/test-genie-mcp.git
cd test-genie-mcp
npm install
npm run build

Running the MCP client via npx is supported once the package is available locally or globally. You can also integrate with Claude Desktop using the provided configuration snippet.

Additional configuration and usage notes

Configure your MCP client to connect to Test Genie MCP using a local stdio command. The client can execute the MCP via npx with the following pattern:

{ 
  "mcpServers": {
    "test-genie": {
      "command": "npx",
      "args": ["test-genie-mcp"]
    }
  }
}

Security and maintenance notes

Maintain your environment by keeping Node.js and npm up to date, reviewing permission settings for global npm installations, and ensuring you have proper access controls for your CI/CD configuration. Regularly update dependencies and monitor AI-generated fix suggestions to verify changes before applying.

Examples and workflows

Example workflow steps you can perform with Test Genie MCP include analyzing the app structure, generating test scenarios, executing tests, detecting issues, suggesting fixes, and generating a final report. Convert these steps into repetitive automation to build a reliable testing loop for ongoing development.

Available tools

analyze_app_structure

Analyze the codebase structure, including screens, components, APIs, and state to understand the app layout.

analyze_code_deep

Perform AST-based code analysis to assess complexity, hooks usage, and dependencies.

analyze_performance

Conduct deep performance analysis covering rendering, computation, network activity, and bundle size.

generate_scenarios

AI-powered generation of test scenarios tailored to the app under test.

create_test_plan

Create test plans and schedules for organized execution.

run_scenario_test

Execute individual scenario tests to validate behavior.

run_simulation

Simulate user behavior with random or pattern-based inputs.

run_stress_test

Perform stress and load testing to evaluate stability under pressure.

detect_memory_leaks

Detect memory leaks through heap analysis and reference tracking.

detect_logic_errors

Identify race conditions and state inconsistencies that affect correctness.

suggest_fixes

AI-powered suggestions for potential code changes to fix issues.

confirm_fix

User confirmation step to approve proposed fixes.

apply_fix

Apply confirmed fixes to the codebase.

rollback_fix

Revert changes if a fix needs to be undone.

run_full_automation

Run the entire automation pipeline from analysis to reporting.

generate_report

Produce detailed reports in Markdown, HTML, or JSON formats.

generate_cicd_config

Create CI/CD configurations for popular platforms.

get_pending_fixes

Retrieve a list of pending fix suggestions awaiting action.

get_test_history

Fetch the history of test executions for auditing.