Python Testing Tools MCP Server

Configuration

The server uses environment variables for configuration:

• GEMINI_API_KEY: Required - Your Google Gemini API key for AI-powered test generation
• GEMINI_MODEL: Optional - Gemini model to use (default: gemini-2.5-flash)

The BAML configuration in baml_src/main.baml defines:

• AI function signatures for test generation, fuzz input creation, and coverage-focused test generation
• Structured response formats using BAML classes (PythonTestFile, FuzzInput, CoverageAnalysis)
• Detailed prompts for comprehensive test coverage, including branch and path analysis

Installation

1. Create and activate a virtual environment:

python3 -m venv .venv
source .venv/bin/activate

2. Install dependencies from requirements.txt:

pip install -r requirements.txt

3. Set your Gemini API key:

export GEMINI_API_KEY="your-api-key-here"

Running the Server

To run the MCP server:

python mcp_server.py

Running the Server in Claude Code

The documentation for FastMCP and Claude Code can be found here: https://gofastmcp.com/integrations/claude-code

Step 1: Install the MCP Server

Install the server using FastMCP:

fastmcp install claude-code python_testing_mcp_server.py --env GEMINI_API_KEY='your-api-key-here'

Step 2: Configure Claude Code

Your .claude.json configuration should look like this:

{
  "/path/to/your/project": {
    "allowedTools": [],
    "history": [],
    "mcpContextUris": [],
    "mcpServers": {
      "python_testing_tools": {
        "type": "stdio",
        "command": "uv",
        "args": [
          "run",
          "--with",
          "fastmcp",
          "fastmcp",
          "run",
          "/path/to/your/project/python_testing_mcp_server.py"
        ],
        "env": {
          "GEMINI_API_KEY": "your-api-key-here"
        }
      }
    },
    "enabledMcpjsonServers": [],
    "disabledMcpjsonServers": [],
    "hasTrustDialogAccepted": false,
    "projectOnboardingSeenCount": 0,
    "hasClaudeMdExternalIncludesApproved": false,
    "hasClaudeMdExternalIncludesWarningShown": false
  }
}

Note: Replace /path/to/your/project with the actual path to your project directory, and your-api-key-here with your actual Gemini API key.

Step 3: Start Claude Code

Start Claude Code:

claude

Step 4: Verify Connection

Check if the MCP server is connected:

/mcp

Step 5: Test the Tools

Test the unit test generation tool:

create unit tests for @demo/basic_example_functions.py

Test the coverage testing tool:

generate comprehensive coverage tests for @demo/basic_example_functions.py

Test the fuzz testing tool:

fuzz test the add function in @demo/basic_example_functions.py

These should create test files in the demo folder with comprehensive test coverage.

Available Tools

Unit Test Generator

• Function: generate_unit_tests_tool
• Description: Generates comprehensive unit tests for all functions in a Python file using AI
• Parameters:
  • file_path (str) - Path to the Python file to generate tests for
• Features:
  • Generates 4-6 test cases per function covering normal, edge, and error cases
  • Uses proper unittest framework with self.assertRaises() for exception testing
  • Automatically handles module imports and class naming conventions
  • Creates properly formatted test files with correct indentation

Fuzz Tester

• Function: fuzz_test_function_tool
• Description: Performs intelligent fuzz testing on a specific function using AI-generated inputs
• Parameters:
  • file_path (str) - Path to the Python file containing the function
  • function_name (str) - Name of the specific function to fuzz test
• Features:
  • Generates 20 diverse test inputs including edge cases and malformed data
  • Tests with boundary values, large inputs, and unusual data combinations
  • Reports crashes with detailed error traces and input values
  • Uses ast.literal_eval() for safe input parsing

Coverage Tester

• Function: generate_coverage_tests_tool
• Description: Generates comprehensive test cases designed to achieve maximum code coverage using advanced AST analysis and AI-powered test generation
• Parameters:
  • file_path (str) - Path to the Python file to generate coverage tests for
• Features:
  • Advanced AST Analysis: Automatically detects branches, loops, exception paths, and return statements
  • Intelligent Coverage Targeting: Generates tests specifically designed to cover all code paths
  • Built-in Coverage Measurement: Integrates coverage.py for real-time coverage reporting
  • Comprehensive Test Scenarios: Creates tests for:
    • Branch coverage (if/elif/else conditions)
    • Loop coverage (zero, one, multiple iterations, break/continue)
    • Exception coverage (try/except/finally blocks)
    • Return path coverage (all possible return statements)
    • Parameter coverage (different types, boundary values, edge cases)
    • Type compatibility testing (invalid inputs, TypeError scenarios)
  • State-of-the-Art Testing: Includes infinity/NaN testing, large number handling, empty collections
  • Automatic Import Resolution: Properly includes all necessary imports (sys, math, etc.)
  • Coverage Reporting: Generates detailed coverage reports showing achieved coverage percentages