network-mcp
Provides AI assistants with direct access to multi-vendor network devices for tasks like configuration management, health checks, and topology discovery through 35 specialized tools. It enables natural language control over platforms including Cisco, Juniper, and Nokia using SSH, NETCONF, and SNMP protocols.
README
network-mcp
Give AI direct access to your network devices. 35 MCP tools, multi-vendor, one interface.
Stop writing Netmiko scripts. Connect Claude, ChatGPT, or any MCP-compatible AI to your routers, switches, and firewalls — and let it run show commands, check health, calculate subnets, discover topology, and manage configs through natural language.
You: "Check the health of all my devices"
Claude: [calls health_check_all] → 6 devices healthy, Switch-R2 has 2 interfaces down
What is this?
An MCP (Model Context Protocol) server that gives AI assistants real-time access to network devices. Built for network engineers who want to automate without writing boilerplate.
Supported platforms:
- Cisco IOS-XE (routers, switches)
- Nokia SR Linux
- FRRouting (FRR)
- Juniper Junos
- Aruba AOS-CX
- Linux hosts
Quick Start
1. Install
git clone https://github.com/E-Conners-Lab/MCP-DEMO-LAB.git
cd network-mcp
python -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt
2. Configure devices
cp .env.example .env
# Edit .env with your device IPs and credentials
3. Try it instantly (no lab required)
# Demo mode returns realistic mock data — no devices needed
DEMO_MODE=true python network_mcp_server.py
4. Connect to Claude Desktop
Add to ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"network": {
"command": "/path/to/network-mcp/.venv/bin/python",
"args": ["/path/to/network-mcp/network_mcp_server.py"]
}
}
}
Restart Claude Desktop. You now have 35 network tools available.
5. Try with a real lab (5 minutes)
# Spin up 2 FRR routers with containerlab
cd quickstart && sudo containerlab deploy -t topology.clab.yml
Tools
Device Operations
| Tool | Description |
|---|---|
get_devices |
List all devices in inventory |
send_command |
Run show commands on any device |
send_config |
Push configuration changes |
health_check |
Check device health (CPU, memory, interfaces) |
health_check_all |
Health check all devices in parallel |
backup_config |
Backup running configuration |
compare_configs |
Diff two config backups |
rollback_config |
Restore a previous config |
Network Intelligence
| Tool | Description |
|---|---|
discover_topology |
LLDP-based topology discovery |
get_routing_table |
View routing tables |
get_neighbors |
BGP/OSPF neighbor status |
get_arp_table |
ARP table lookup |
get_mac_table |
MAC address table |
ping_sweep |
Sweep a subnet for reachable hosts |
traceroute |
Trace path to destination |
Calculators (no devices needed)
| Tool | Description |
|---|---|
calculate_tunnel_mtu |
Optimal MTU/MSS for VPN tunnels |
calculate_subnet_info |
Subnet details from CIDR notation |
split_network |
VLSM subnet splitting |
convert_netmask |
CIDR to dotted decimal conversion |
SNMP & Monitoring
| Tool | Description |
|---|---|
snmp_get_oid |
SNMP GET for specific OIDs |
snmp_walk_oid |
SNMP WALK subtrees |
snmp_poll_metrics |
Poll interface/CPU/memory metrics |
NETCONF
| Tool | Description |
|---|---|
get_interfaces_netconf |
Interface data via NETCONF |
get_bgp_neighbors_netconf |
BGP state via NETCONF |
get_netconf_capabilities |
Device NETCONF capabilities |
Configuration Management
| Tool | Description |
|---|---|
compliance_check |
Check device against compliance templates |
full_network_test |
End-to-end network validation |
Architecture
┌─────────────────────────────────────────┐
│ AI Assistant (Claude, ChatGPT, etc.) │
└────────────────┬────────────────────────┘
│ MCP Protocol (stdio/SSE)
┌────────────────▼────────────────────────┐
│ network_mcp_server.py │
│ FastMCP server + tool registry │
└────────────────┬────────────────────────┘
│
┌────────────────▼────────────────────────┐
│ mcp_tools/ │
│ ├── device.py (9 tools) │
│ ├── calculators.py (6 tools) │
│ ├── topology.py (6 tools) │
│ ├── config.py (8 tools) │
│ ├── snmp.py (5 tools) │
│ ├── netconf.py (4 tools) │
│ ├── compliance.py (7 tools) │
│ └── ... (10 modules, 35 tools total) │
└────────────────┬────────────────────────┘
│
┌────────────────▼────────────────────────┐
│ core/ │
│ ├── connection.py (Netmiko/Scrapli) │
│ ├── parser.py (NTC/Genie) │
│ ├── mtu_calculator.py │
│ └── subnet_calculator.py │
└────────────────┬────────────────────────┘
│ SSH / NETCONF / SNMP
┌────────────────▼────────────────────────┐
│ Network Devices │
│ Cisco · Nokia · FRR · Juniper · Aruba │
└─────────────────────────────────────────┘
Project Structure
network-mcp/
├── network_mcp_server.py # MCP server entry point
├── config/
│ └── devices.py # Device inventory (env var overrides)
├── mcp_tools/ # All MCP tool implementations
│ ├── device.py # Core device operations
│ ├── calculators.py # MTU/subnet calculators
│ ├── topology.py # LLDP discovery
│ ├── config.py # Config backup/compare/rollback
│ ├── snmp.py # SNMP polling
│ ├── netconf.py # NETCONF operations
│ ├── compliance.py # Compliance checking
│ └── ... # 10 modules, 35 tools total
├── core/ # Connection and parsing libraries
├── templates/ # Jinja2 config templates (FRR, IOS-XE)
├── quickstart/ # Containerlab quick-start lab
└── tests/ # Test suite
Configuration
All configuration is via environment variables (or .env file):
# Device credentials
DEVICE_USERNAME=admin
DEVICE_PASSWORD=admin
# Device IPs (override defaults)
R1_HOST=10.255.255.11
R2_HOST=10.255.255.12
# Optional features
DEMO_MODE=true # Mock data, no real devices
USE_NETBOX=true # Pull inventory from NetBox
NETBOX_URL=http://localhost:8000
NETBOX_TOKEN=your-token
Requirements
- Python 3.11+
- Network devices reachable via SSH/NETCONF/SNMP
- Claude Desktop or any MCP-compatible client
Contributing
See CONTRIBUTING.md for guidelines.
License
MIT License - see LICENSE for details.
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.