Skip to content

Prerequisites

Quick checklist before installing MCP Mesh 0.3.x with dual decorator pattern

Essential Requirements for MCP Mesh 0.3.x

You need Python 3.9+ and understanding of the new dual decorator pattern:

1. Python 3.9+

# Check your Python version
python --version

# Should show: Python 3.9.x or higher

Don't have Python 3.9+?

  • macOS: brew install python@3.11
  • Ubuntu/Debian: sudo apt install python3.11
  • Windows: Download from python.org

2. pip (Python Package Manager)

# Check pip is installed
pip --version

# If not installed, get it with:
python -m ensurepip --upgrade

3. Understanding the Dual Decorator Pattern

MCP Mesh 0.3.x introduces the dual decorator pattern combining:

  • FastMCP decorators (@app.tool, @app.prompt, @app.resource) - Familiar MCP development
  • Mesh decorators (@mesh.tool, @mesh.agent) - Orchestration and dependency injection
import mesh
from fastmcp import FastMCP

app = FastMCP("My Service")

@app.tool()      # ← FastMCP decorator
@mesh.tool(      # ← Mesh decorator
    capability="greeting",
    dependencies=["time_service"]
)
def hello(time_service: mesh.McpMeshAgent = None):
    return f"Hello! Time: {time_service()}"

That's it! You're ready to install MCP Mesh 0.3.x.

For the best experience, we recommend:

Use a Virtual Environment

# Create a virtual environment
python -m venv mcp-env

# Activate it
source mcp-env/bin/activate  # Linux/macOS
# or
mcp-env\Scripts\activate     # Windows

Have curl or wget

# For testing your agents
curl --version

System Support

  • ✅ Linux: All distributions with Python 3.9+
  • ✅ macOS: 10.15 (Catalina) or later
  • ✅ Windows: Windows 10/11 (WSL2 recommended for best experience)

Quick Check Script

# Run this to check everything at once
python3 -c "
import sys
print('Python:', sys.version)
print('✅ Ready!' if sys.version_info >= (3, 9) else '❌ Need Python 3.9+')
"

Optional but Helpful

Network Ports

MCP Mesh will use these ports (configurable):

  • 8000: Registry (starts automatically)
  • 8080-8090: Your agents (you choose)

Storage

  • 500MB: For MCP Mesh and dependencies
  • 100MB: For logs and data

Next Steps

Once all prerequisites are met, proceed to Installation →

💡 Tip: If you encounter issues, our Troubleshooting Guide covers common problems and solutions.

📚 Note: For containerized deployments (Docker/Kubernetes), different prerequisites apply. See Docker Deployment or Kubernetes Basics.

🔧 Troubleshooting

Common Issues

  1. Python version mismatch - Use pyenv or conda to manage multiple Python versions
  2. Permission denied on ports - Either use higher ports (>1024) or run with appropriate permissions
  3. Git not found - Install via package manager (apt, brew, choco)
  4. Virtual environment issues - Ensure you're using the Python 3 venv module, not virtualenv

For detailed solutions, see our Troubleshooting Guide.

⚠️ Known Limitations

  • Python 3.8 and below: Not supported due to typing features used
  • 32-bit systems: Limited support, 64-bit recommended
  • Network proxies: May require additional configuration
  • Corporate firewalls: May block agent communication on custom ports

📝 TODO

  • Add automated prerequisite installer script
  • Support for Python 3.13 when released
  • Add Podman as Docker alternative
  • Create offline installation package
  • Add ARM64 native support verification