Why MCP Mesh?¶

The Problem: Building Agents is Easy, Deploying Them is Hard¶

Every AI framework lets you build agents. But when it's time to deploy:

• How do you deploy 10 agents to Kubernetes?
• How do agents discover each other at runtime?
• How do you scale Agent A independently from Agent B?
• How do you monitor which agent called which?

The answer from most frameworks: "That's your problem."

MCP Mesh: Build AND Deploy¶

MCP Mesh is a complete platform for building and deploying AI agents. You don't need LangGraph, CrewAI, or AutoGen.

What MCP Mesh Provides¶

Simple Example¶

from mcp_mesh import mesh
from fastmcp import FastMCP

app = FastMCP("My Agent")

# Define a tool with automatic dependency injection
@app.tool()
@mesh.tool(capability="greeting", dependencies=["time_service"])
async def greet(name: str, time_service: mesh.McpMeshAgent = None):
    current_time = await time_service() if time_service else "unknown"
    return f"Hello {name}! The time is {current_time}"

# Expose a REST endpoint
@mesh.route("/api/greet/{name}", methods=["GET"])
async def greet_api(name: str):
    return {"message": greet(name)}

# Register as an agent
@mesh.agent(name="greeting-agent", port=8080)
class GreetingAgent:
    pass

That's a complete agent with:

• Tool that other agents can call
• REST API for external access
• Automatic registration with the mesh
• Dependency injection from other agents

Framework Comparison¶

Key Insight¶

• LangGraph/CrewAI/AutoGen = Agent building frameworks (no deployment)
• MCP Mesh = Agent building + deployment + scaling + observability

What Makes MCP Mesh Different¶

1. Decorators That Do Everything¶

@mesh.agent(name="my-agent", port=8080)      # Register with mesh
@mesh.tool(capability="process")              # Expose as callable tool
@mesh.llm(provider="anthropic")               # Add LLM capabilities
@mesh.route("/api/endpoint")                  # REST API endpoint

2. Automatic Dependency Injection¶

@mesh.tool(
    capability="analyze_data",
    dependencies=["db_service", "ml_service"]
)
async def analyze(data, db_service: mesh.McpMeshAgent = None, ml_service: mesh.McpMeshAgent = None):
    # db_service and ml_service are automatically injected
    # MCP Mesh finds them, connects them, handles failures
    records = await db_service(query=data) if db_service else {}
    return await ml_service(data=records) if ml_service else {}

3. One Command to Production¶

# Generate Docker Compose with observability
meshctl scaffold --compose --observability

# Or deploy to Kubernetes (OCI registry)
helm install my-mesh oci://ghcr.io/dhyansraj/mcp-mesh/mcp-mesh-core \
  --version 0.7.21 -n mcp-mesh --create-namespace

4. Built-in Observability¶

• Grafana dashboards - Pre-configured for agent metrics
• Distributed tracing - See request flow across agents
• Health monitoring - Automatic health checks and alerting

When to Use MCP Mesh¶

Use MCP Mesh When:¶

• Building multi-agent AI systems
• Deploying agents to Kubernetes or Docker
• Agents need to discover and call each other
• You want independent scaling per agent
• You need production observability

Maybe Skip MCP Mesh When:¶

• Single agent running locally only
• Pure prototyping with no deployment plans

Quick Comparison: Deploying 5 Agents¶

Without MCP Mesh¶

├── agent-1/
│   ├── Dockerfile
│   ├── deployment.yaml
│   ├── service.yaml
│   └── configmap.yaml
├── agent-2/
│   └── ... (repeat)
├── service-discovery/
│   └── (build your own)
└── monitoring/
    └── (set up yourself)

Result: 50+ files, weeks of work

With MCP Mesh¶

# 5 Python files with @mesh.agent decorators

meshctl scaffold --compose --observability
docker-compose up

Result: 5 files + 1 command

Already Using Another Framework?¶

If you're already invested in LangGraph, CrewAI, or AutoGen, MCP Mesh can help you deploy them to production:

• Wrap your existing agents with @mesh.agent
• Get automatic K8s deployment via Helm
• Add observability without code changes
• Scale agents independently

MCP Mesh doesn't replace your agent logic - it handles the infrastructure so you don't have to.

Next Steps¶

Ready to build your first agent?

Get Started in 5 Minutes View Architecture