gitlab-testplan-generator

Atm06/gitlab-testplan-generator

3.1

If you are the rightful owner of gitlab-testplan-generator and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

An AI-powered MCP server that generates comprehensive UI test plans from GitLab merge requests using local Ollama AI models. Your code never leaves your machine.

Tools
1
Resources
0
Prompts
0

GitLab UI Test Plan Generator

An AI-powered MCP server that generates comprehensive UI test plans from GitLab merge requests using local Ollama AI models. Your code never leaves your machine.

🏗️ Project Structure

gitlab-testplan-generator/
├── mcp-server/
│   ├── src/gitlab_mcp/           # Main package
│   │   ├── server.py            # MCP server with ui_test_plan_from_mr tool
│   │   ├── analyzer.py          # GitLab MR analysis with AI
│   │   ├── test_planner.py      # AI-powered test plan generation
│   │   ├── ai_service.py        # Local Ollama AI integration
│   │   └── models.py            # Pydantic data models
│   ├── examples/                # Usage examples
│   │   ├── demo_analysis.py     # Demo script
│   │   ├── example_usage.py     # Example usage patterns
│   │   └── test_ollama.py       # AI agent verification test
│   ├── docs/                    # Documentation
│   ├── main.py                  # Entry point
│   ├── requirements.txt         # Dependencies
│   └── .env                     # Environment configuration
└── venv/                        # Virtual environment

🔧 Features

🤖 AI-Powered Analysis

  • Uses local Ollama AI models for intelligent code analysis
  • Analyzes GitLab merge requests to identify affected UI areas
  • Generates context-aware test scenarios based on code changes
  • Privacy-first: All analysis happens locally on your machine

📋 Comprehensive Test Plans

  • Generates UI-focused test scenarios with specific steps
  • Risk assessment for each test scenario (high/medium/low)
  • Expected results and acceptance criteria
  • Handles both positive and edge-case testing scenarios

🔗 Seamless Integration

  • Direct integration with Cursor IDE via MCP protocol
  • Simple @ui_test_plan_from_mr command usage
  • Works with GitLab.com and self-hosted GitLab instances
  • No external API calls - everything runs locally

🚀 Quick Start

1. Prerequisites

  • Python 3.8+
  • Ollama installed and running
  • GitLab personal access token
  • Cursor IDE

2. Installation

# Navigate to the project
cd gitlab-testplan-generator

# Create and activate virtual environment
python3 -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install dependencies
cd mcp-server
pip install -r requirements.txt

3. Configuration

Set up your GitLab credentials in mcp-server/.env:

GITLAB_URL=https://gitlab.cee.redhat.com  # Your GitLab instance
GITLAB_TOKEN=your-personal-access-token   # GitLab PAT with api, read_repository scopes
GITLAB_SSL_VERIFY=false                   # For internal GitLab instances

4. Install in Cursor

Add to your ~/.cursor/mcp.json:

{
  "mcpServers": {
    "gitlab-ui-test-plan-generator": {
      "command": "/path/to/gitlab-testplan-generator/venv/bin/python",
      "args": ["/path/to/gitlab-testplan-generator/mcp-server/main.py"],
      "env": {
        "GITLAB_URL": "https://gitlab.cee.redhat.com"
      }
    }
  }
}


### 5. Start Ollama

```bash
# Install and start Ollama
ollama serve

# Pull the AI model (in another terminal)
ollama pull qwen2.5-coder:1.5b

6. Verify AI Agent (Optional but Recommended)

# Navigate to examples directory
cd mcp-server/examples

# Run AI agent verification test
python test_ollama.py

This test verifies:

  • ✅ Ollama server status and version
  • ✅ Required AI model availability
  • ✅ AI generation capabilities

Expected output:

=== Ollama AI Agent Status Check ===
✅ Ollama server is running (version: 0.9.6)
✅ Required model 'qwen2.5-coder:1.5b' is available
✅ AI generation successful!
🎉 All tests passed! Your AI agent is running correctly on Ollama.

Demo & Examples

# Verify AI agent is working correctly
python examples/test_ollama.py

# Run demo analysis (works offline without GitLab credentials)
python examples/demo_analysis.py

# Run with real GitLab connection
# (requires configured .env file)
python examples/demo_analysis.py

📋 Available MCP Tool

ui_test_plan_from_mr

The main tool for generating UI test plans from GitLab merge requests.

Usage in Cursor:

@ui_test_plan_from_mr https://gitlab.cee.redhat.com/project/-/merge_requests/123

What it does:

  1. 🔍 Analyzes the merge request and extracts code changes
  2. 🤖 Uses local AI to understand the impact of changes
  3. 🎯 Identifies affected UI areas and components
  4. 📋 Generates comprehensive test scenarios with specific steps
  5. ⚠️ Provides risk assessment for each scenario

💡 Usage Examples

In Cursor:

@ui_test_plan_from_mr https://gitlab.cee.redhat.com/<project_name>/merge_requests/12345

@ui_test_plan_from_mr https://gitlab.cee.redhat.com/project/-/merge_requests/123/diffs

Example Output:

The tool generates:

  • MR Title: [JIRA-ID]: MR title description
  • Affected Pages: Affected pages and components
  • Test Scenarios: 2-5 AI-generated scenarios
  • Risk Levels: High/Medium/Low for each scenario
  • Detailed Steps: Specific actions and expected results
  • AI Insights: Summary of changes and user impact

🔧 Architecture

Core Components

  • CodeAnalyzer: Analyzes GitLab merge requests and calculates complexity/risk scores
  • UITestPlanGenerator: Creates comprehensive test plans based on change analysis
  • MCP Server: Exposes functionality through the Model Context Protocol
  • Models: Pydantic data models for type safety and validation

Component Analysis

The system automatically identifies components based on file patterns:

  • Authentication: /auth/, /login/, /user/ files
  • Database: .sql, migration files, /models/, /db/ files
  • API: /api/, /endpoints/, /routes/ files
  • Frontend: .js, .ts, .vue, .html, .css files
  • Backend: .py, .java, .cpp, .go files
  • Configuration: /config/, .env, .yaml, .json files
  • Security: /security/, /ssl/ files

Risk Assessment

The server automatically assesses risk levels:

  • High Risk: Security files, database migrations, configuration files, complex changes (>100 lines)
  • Medium Risk: API files, service files, moderate complexity (50-100 lines)
  • Low Risk: Documentation, simple changes (<50 lines)

📖 Documentation

  • : Detailed setup instructions
  • : Working code examples and verification scripts

🤝 Contributing

  1. Fork the repository
  2. Create a feature branch: git checkout -b feature-name
  3. Make your changes and test manually
  4. Verify AI agent functionality: python examples/test_ollama.py
  5. Test with real GitLab merge requests
  6. Submit a pull request

Development Setup

# Install dependencies
pip install -r requirements.txt

# Verify setup
python examples/demo_analysis.py

📝 License

MIT License - see LICENSE file for details.

🆘 Troubleshooting

Common Issues

  1. Import Errors: Make sure you've installed the package with pip install -e .
  2. GitLab Connection: Verify your token has correct permissions and hasn't expired
  3. MCP Installation: Ensure Cursor has MCP support enabled

Debug Mode

# Run with debug logging
export PYTHONPATH=src
python -m gitlab_mcp.server --log-level debug

Ready to start? Check out the for detailed instructions!