#mcpserver #mcptutorial #mcp #Cisco #NetworkAutomation #devnet #AIfornetworking #networkengineer
Join this channel to get access to perks:
https://www.youtube.com/channel/UCcA2nhdC0wzqyv9x1lk5NnA/join
📚 Related Playlists
🔹 MCP Protocol Tutorial Playlist → https://www.youtube.com/watch?v=G9Ct5Dl4D3k&list=PLOocymQm7YWZuDHVYSvF513AxTfMs_tFC
🔹 Generative AI Tutorial Full Videos Playlist → https://www.youtube.com/watch?v=32Z7r8LcY8o&list=PLOocymQm7YWYODnubEmdLTfuj-T7EpOuw
Welcome to the second video of the Model Context Protocol (MCP) Demo Series! 🎉
In this video, we dive deep into building a Network Device Controller MCP Server that allows you to interact with Cisco network devices directly from Visual Studio Code (VS Code), GitHub Copilot, and Cloud Desktop. If you’re a network engineer, DevNet enthusiast, or automation architect, this is a hands-on guide to integrating AI-driven workflows with network automation.
In the first video of this series, we created a simple To-Do List MCP server and integrated it with GitHub Copilot and Cloud Desktop. Now, we take a big step forward by creating a real-world MCP server that connects to Cisco IOS-XR and IOS-XE devices. You’ll learn how to run operational commands, retrieve interface details, assign IP addresses to loopback interfaces, and even delete interfaces — all automated through MCP tools.
🔹 What This Video Covers in Detail:
1️⃣ Introduction to MCP for Networking
Quick recap of Part 1 (To-Do list MCP server).
Use case for extending MCP to network device management.
Why MCP matters for AI-assisted network automation workflows.
2️⃣ Building a Network Device Controller MCP Server
Creating a new project with uv (Python project manager).
Setting up a proper directory structure and files.
Writing a device inventory file (device_inventory.py) to manage multiple devices (CSR router, Cisco DevNet device, IOS-XR, IOS-XE).
Storing credentials, device type, and connection details for XR vs XE.
3️⃣ Implementing Core Functions in Python
Writing get_credentials() to fetch device login details from inventory.
Creating connect_and_run() for running show commands on Cisco devices.
Installing and configuring Netmiko for secure device access.
Using send_command and send_config_set methods for read/write operations.
4️⃣ Executing Cisco Commands with MCP
Running show version, show ip interface brief, and SNMP-specific configs.
Handling differences between Cisco IOS-XR (commit) and IOS-XE (write memory).
Automating execution of commands using MCP tools.
5️⃣ Configuring and Managing Interfaces
Creating connect_and_config() to push configuration changes.
Assigning IP addresses to loopback interfaces dynamically.
Auto-calculating subnet masks for /30 networks.
Deleting loopback interfaces from XR and XE devices with commit handling.
6️⃣ Testing with MCP Inspect and Copilot
Using uv run mcp-dev controller.py to inspect and test MCP tools.
Running commands interactively with MCP Inspect.
Adding the Network Controller MCP server to VS Code settings.json.
Using GitHub Copilot in Agent Mode to trigger MCP functions:
Example: “Get the interface status of R2”
Example: “Show SNMP server config of R1 and explain each line”
7️⃣ Integration with Cloud Desktop
Adding the MCP server path in Cloud Desktop config.
Running commands via Cloud Desktop tools.
Understanding token usage limitations in Cloud Desktop vs Copilot.
8️⃣ Real-World Benefits of MCP for Networking
Unified automation across multiple platforms (VS Code, Copilot, Cloud Desktop).
Standardized approach for network automation and AI workflows.
Easily extendable MCP server to support more commands (e.g., route details, ACLs, VLANs).
💡 Why This Matters
Traditionally, network engineers manually log into Cisco devices, type commands, copy outputs, and push configs. With this approach, you can:
✔️ Automate repetitive tasks like checking interface status.
✔️ Standardize config changes (IP address assignment, loopback management).
✔️ Reduce errors by letting MCP + AI (Copilot/Cloud Desktop) generate and execute commands.
✔️ Extend MCP servers with custom tools tailored to your environment.
This video gives you practical, step-by-step knowledge you can apply in real labs or production networks. Whether you’re exploring Cisco DevNet certification, network automation frameworks, or AI-driven workflows, this tutorial will help you bridge the gap between Python scripting and MCP-based automation.
