ZAP MCP Server
Integrates OWASP ZAP security testing with AI assistants through MCP, enabling automated vulnerability scanning and AI-powered security analysis during development. Supports multiple scan types including active, passive, and AJAX spider scans with real-time status updates.
README
ZAP MCP Server
A powerful Model Context Protocol (MCP) Server that integrates OWASP ZAP (Zed Attack Proxy) with AI assistants and MCP clients. Enable AI-powered security testing through automated vulnerability scanning.
🎯 Why ZAP MCP Server?
Shift Left Security - Empower developers to integrate security testing early in the development lifecycle. Instead of waiting for security reviews at the end of development, developers can now:
- 🔧 Test during development - Run security scans on localhost applications
- 🤖 AI-assisted security - Get intelligent vulnerability analysis through AI assistants
- ⚡ Rapid feedback - Identify security issues before they reach production
- 🔄 CI/CD integration - Automate security testing in development workflows
- 📊 Developer-friendly - Simple MCP interface for non-security experts
🚀 Features
- 🔍 Multiple Scan Types: Active, Passive, AJAX Spider, and Complete scans
- ⚡ Asynchronous Processing: Background scan execution with real-time status updates
- 🐳 Docker Support: Easy deployment with Docker Compose
- 🤖 AI Integration: Seamless integration with MCP-compatible AI assistants
- 📊 Rich Reporting: Detailed vulnerability reports with risk scoring
- 🔄 Session Management: Flexible session handling strategies
- 🛡️ Production Ready: Robust error handling and logging
- 🔄 Automatic URL Transformation: Automatically maps
localhostURLs to container host gateways
📋 Prerequisites
- Python 3.8+
- OWASP ZAP installed and accessible via PATH
- Java (required by ZAP)
- Docker or Podman (optional, for containerized deployment)
- Firefox ESR (included in Docker/Podman containers for AJAX scans)
📖 For container-specific prerequisites, see:
🛠️ Installation
Package Structure
This project uses a proper Python package structure (zap_custom_mcp/) which provides several benefits:
- ✅ Clean imports - Proper module organization
- ✅ Docker compatibility - Works seamlessly in containers
- ✅ PyPI ready - Can be published as a proper Python package
Execution methods:
python -m zap_custom_mcp(recommended)python -m zap_custom_mcp.http_server(alternative method)
Option 1: Local Installation
-
Clone the repository
git clone https://github.com/LisBerndt/zap-custom-mcp.git cd zap-custom-mcp -
Install OWASP ZAP
- Download from OWASP ZAP Downloads
- Ensure
zap.batis accessible via PATH - Test:
where zap.bat(Windows) orwhich zap.sh(Linux/Mac)
-
Install Python dependencies
pip install -r requirements.txt -
Linux-specific notes
-
Install Java (OpenJDK 11+ recommended):
# Debian/Ubuntu sudo apt-get update && sudo apt-get install -y default-jre # Fedora sudo dnf install -y java-11-openjdk -
Install OWASP ZAP:
# Debian/Ubuntu (from official repos) sudo apt-get update && sudo apt-get install -y zaproxy # If not available in your distro repos, download from ZAP website: # https://www.zaproxy.org/download/ -
Verify ZAP is on PATH (Linux/Mac):
which zap.sh || echo "zap.sh not found in PATH"
-
Option 2: Docker/Podman Deployment (Recommended)
🐳 Docker/Podman is the easiest and most reliable method!
# 1. Clone repository
git clone https://github.com/LisBerndt/zap-custom-mcp.git
cd zap-custom-mcp
# 2. Build and start containers (auto-detects Docker/Podman)
# Linux/Mac:
./build.sh
./start.sh
# Windows:
build.bat
start.bat
# 3. Check status
docker-compose ps # or podman-compose ps
📖 For detailed Docker/Podman setup and localhost scanning instructions, see:
✅ AUTOMATIC URL TRANSFORMATION: The server automatically detects Docker/Podman environments and transforms localhost/127.0.0.1 URLs to the appropriate host gateway:
- Docker:
localhost:3000→host.docker.internal:3000 - Podman:
localhost:3000→host.containers.internal:3000 - Local: URLs remain unchanged
This means you can use localhost URLs directly - they will be automatically transformed when running in containers!
⚙️ Configuration
The server uses environment variables for configuration. Key settings:
| Variable | Default | Description |
|---|---|---|
ZAP_BASE |
http://127.0.0.1:8080 |
ZAP API port - Change port by modifying URL |
ZAP_MCP_PORT |
8082 |
MCP server port - Port for MCP client connections |
ZAP_MCP_HOST |
127.0.0.1 |
MCP server host (use 0.0.0.0 for all interfaces) |
ZAP_AUTOSTART |
true |
Auto-start ZAP if not running |
ZAP_LOG_LEVEL |
INFO |
Logging level |
Custom Port Configuration
Using .env file (Recommended):
# Copy the example file
cp env.example .env
# Edit .env file
ZAP_PORT=8081
ZAP_MCP_PORT=8083
ZAP_BASE=http://127.0.0.1:8081
Using environment variables:
# Set custom ports
export ZAP_PORT=8081
export ZAP_MCP_PORT=8083
export ZAP_BASE="http://127.0.0.1:8081"
# Then start containers
./start.sh
📖 For complete configuration details, see:
🚀 Quick Start
🐳 Docker/Podman (Recommended)
Fastest start with containers:
# 1. Clone repository
git clone https://github.com/LisBerndt/zap-custom-mcp.git
cd zap-custom-mcp
# 2. Start (auto-detects Docker/Podman)
# Linux/Mac:
./build.sh && ./start.sh
# Windows:
build.bat
start.bat
# 3. Wait (approx. 90 seconds) and then connect:
# http://localhost:8082/mcp
📖 For detailed setup instructions and localhost scanning, see:
✅ AUTOMATIC URL TRANSFORMATION: The server automatically detects Docker/Podman environments and transforms localhost/127.0.0.1 URLs to the appropriate host gateway. You can use localhost URLs directly - no manual mapping needed!
💻 Local Installation
1. Start the Server
Recommended method (as package):
python -m zap_custom_mcp
Alternative methods:
# As specific module
python -m zap_custom_mcp.http_server
# Direct execution (legacy, may have import issues)
python zap_custom_mcp/http_server.py
💡 Best Practice: Always use python -m zap_custom_mcp for the most reliable execution.
The server will automatically:
- ✅ Check if ZAP is running
- ✅ Start ZAP if needed (via PATH)
- ✅ Create/load a session
- ✅ Start the MCP server
⏱️ Important: The server takes approximately 90 seconds to become fully operational after startup. This includes:
- ZAP initialization and startup
- Session creation
- MCP server initialization
- All components becoming ready
2. Connect Your MCP Client
Connect to: http://localhost:8082/mcp
MCP Configuration Example
For Cursor IDE, add to your mcp.json:
{
"mcpServers": {
"zap-mcp": {
"url": "http://localhost:8082/mcp"
}
}
}
For other MCP clients, use the same URL endpoint.
3. Available Tools
| Tool | Description |
|---|---|
start_active_scan |
Run active security scan (Spider + Active) |
start_complete_scan |
Run complete scan (AJAX + Spider + Active + Passive) |
start_passive_scan |
Run passive security analysis |
start_ajax_scan |
Run AJAX spider for modern web apps |
get_scan_status |
Get real-time scan status |
cancel_scan |
Cancel running scan |
list_scans |
List all active scans |
create_new_session |
Create new ZAP session |
📖 Usage Examples
Development Workflow Integration
Local Development Testing:
{
"tool": "start_passive_scan",
"arguments": {
"url": "http://localhost:3000",
"timeout_seconds": 60
}
}
Pre-Commit Security Check:
{
"tool": "start_active_scan",
"arguments": {
"url": "http://localhost:8080",
"ascan_max_wait_seconds": 300,
"spider_max_wait_seconds": 120
}
}
Localhost Scanning Examples
✅ AUTOMATIC URL TRANSFORMATION: The server automatically detects Docker/Podman environments and transforms localhost/127.0.0.1 URLs:
{
"tool": "start_complete_scan",
"arguments": {
"url": "http://localhost:3000",
"include_findings": true
}
}
What happens automatically:
- Docker:
http://localhost:3000→http://host.docker.internal:3000 - Podman:
http://localhost:3000→http://host.containers.internal:3000 - Local:
http://localhost:3000→http://localhost:3000(unchanged)
📖 For Docker/Podman localhost scanning examples, see:
Basic Security Scan
💡 Example Target: OWASP Juice Shop - A deliberately vulnerable web application designed for security testing and training.
{
"tool": "start_complete_scan",
"arguments": {
"url": "https://juice-shop.herokuapp.com/#/",
"include_findings": true,
"include_evidence": false
}
}
Quick Passive Scan
💡 Perfect for: Quick security assessment without active testing.
{
"tool": "start_passive_scan",
"arguments": {
"url": "https://juice-shop.herokuapp.com/#/",
"timeout_seconds": 300
}
}
AJAX Spider Scan
💡 Perfect for: Modern web applications with JavaScript/AJAX content.
{
"tool": "start_ajax_scan",
"arguments": {
"url": "https://juice-shop.herokuapp.com/#/",
"maxDuration": 5,
"maxCrawlDepth": 5,
"numberOfBrowsers": 1,
"browserId": "firefox-headless"
}
}
Note: AJAX scans require Firefox to be installed in the container (included by default). Firefox runs in headless mode without requiring any display server.
Custom Active Scan
💡 Advanced configuration: Custom timeouts and scan policies for thorough testing.
{
"tool": "start_active_scan",
"arguments": {
"url": "https://juice-shop.herokuapp.com/#/",
"ascan_max_wait_seconds": 3600,
"spider_max_wait_seconds": 900,
"scanPolicyName": "Default Policy"
}
}
🔄 Shift Left Security Integration
Development Workflows
1. Local Development
- Test your localhost application during development
- Get immediate feedback on security issues
- Fix vulnerabilities before committing code
2. Pre-Commit Hooks
- Integrate security scans into git pre-commit hooks
- Prevent insecure code from entering the repository
- Automated security validation
3. CI/CD Pipeline Integration
- Add security testing to your build pipeline
- Scan staging environments automatically
- Generate security reports for each deployment
4. AI-Assisted Security
- Use AI assistants to interpret scan results
- Get recommendations for fixing vulnerabilities
- Learn security best practices through AI guidance
Benefits for Development Teams
- ⚡ Faster feedback - Catch issues in minutes, not weeks
- 💰 Cost reduction - Fix issues early when they're cheaper to resolve
- 🎯 Developer education - Learn security through hands-on testing
- 🛡️ Proactive security - Build secure applications from the start
- 📊 Continuous improvement - Regular security assessments
🐳 Container Deployment
✅ AUTOMATIC URL TRANSFORMATION: The server automatically detects Docker/Podman environments and transforms localhost/127.0.0.1 URLs to the appropriate host gateway. You can use localhost URLs directly!
📖 For complete container setup and usage instructions, see:
- DOCKER.md - Complete Docker setup guide with automatic URL transformation
- PODMAN.md - Complete Podman setup guide with automatic URL transformation
Quick container commands:
# Start containers (auto-detects Docker/Podman)
# Linux/Mac:
./start.sh
# Windows:
start.bat
# Follow logs
docker-compose logs -f # Docker
podman-compose logs -f # Podman
# Stop containers
docker-compose down # Docker
podman-compose down # Podman
📊 Scan Results
Scans return structured results including:
{
"scan_id": "abc12345",
"target": "https://juice-shop.herokuapp.com/#/",
"alerts": {
"High": 2,
"Medium": 5,
"Low": 12,
"Informational": 8
},
"totalAlerts": 27,
"riskScore": 31,
"vulnerabilityNames": [
{ "name": "SQL Injection", "risk": "High", "count": 1 },
{ "name": "XSS", "risk": "Medium", "count": 3 }
],
"durations": {
"ajax": 45.2,
"spider": 120.5,
"ascan": 1800.0,
"pscan": 30.1
}
}
🔧 Troubleshooting
Server Takes Too Long to Start
The server requires approximately 90 seconds to become fully operational. This is normal and includes:
- ZAP startup and initialization
- Session creation
- MCP server initialization
Wait for the startup process to complete before attempting to connect.
ZAP Won't Start
# Check if ZAP is in PATH
where zap.bat # Windows
which zap.sh # Linux/Mac
# Check Java installation
java -version
# Enable debug logging
set ZAP_LOG_LEVEL=DEBUG
python -m zap_custom_mcp
Connection Issues
# Check if ZAP is running
curl http://localhost:8080/JSON/core/view/version/
# Check MCP server
curl http://localhost:8082/mcp
# Check firewall settings
MCP Client Connection Issues
If your MCP client cannot connect:
- Ensure the server has been running for at least 90 seconds
- Verify the URL is correct:
http://localhost:8082/mcp - Check that no firewall is blocking port 8082
- For Cursor IDE, ensure your
mcp.jsonconfiguration is correct
Container Issues
📖 For detailed container troubleshooting, see:
🤝 Contributing
- Fork the repository
- Create a feature branch:
git checkout -b feature-name - Install development dependencies:
pip install -e ".[dev]" - Run tests:
pytest - Commit changes:
git commit -am 'Add feature' - Push to branch:
git push origin feature-name - Submit a Pull Request
Development Setup
# Clone and setup
git clone https://github.com/LisBerndt/zap-custom-mcp.git
cd zap-custom-mcp
# Install in development mode
pip install -e ".[dev]"
# Run tests
pytest
# Start development server
python -m zap_custom_mcp
📄 License
This project is licensed under the MIT License - see the LICENSE file for details.
🙏 Acknowledgments
- OWASP ZAP - The amazing security testing tool
- Model Context Protocol - The protocol that makes AI integration possible
- Sovereign Engineering - SEC-05 cohort for inspiring freedom tech and self-sovereign applications
Special Thanks
This project was inspired by the Sovereign Engineering Community and their commitment to building tools for a self-sovereign future. The SEC-05 cohort's dedication to freedom tech, censorship resistance, and permissionless access aligns perfectly with the goals of making security testing tools more accessible and decentralized.
"Build applications and services for a self-sovereign future." — Sovereign Engineering
📞 Support
Vibe coded with ❤️ for the self sovereign engineer - YOLO!
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.