The model contextual protocol servers provide a new approach to unifying automation and visibility across Cisco hybrid environments. They enable the AI customer to automatically discover and use tools across several catalyst clusters and Meraki organizations.
If you are interested in how it works, now is time to see in action.
In this new demo, Cisco principle technical marketing engineer Gabi NPP It shows how the only AI customer directs questions about the natural language to the right tool, loads resorts from multiple domans, and helps you remove or report more effectively about your network.
See MCP in Action: Catalyst and Meraki Integration
In the video below, Gabi shows how MCP servers allow AI AI to interact with multiple platforms. You will find out:
- How the client connects to multiple MCP servers and discovers available tools.
- How these tools are selected and performed in real time based on the user’s intention.
- How one question can bridge clusters and organizations using patterns like cluster = all.
The video included practical passages of Multi-Klava’s inventory search, correlation of publication and BGP work problems created from basic tools.
Understanding the architecture of MCP and the workflow
MCP customers to Customer-Server, which allows AI assistant to connect to multiple MCP servers and dynamically discover available tool definitions. Here’s what a complete workflow looks like:
- The AI, powered by a large language model, connects to multiple MCP servers.
- Each server provides a list of tools-a built-in Runbus or API-generated API.
- Asks the question; The AI client selects a suitable tool, fills in the parameters and sends the request.
- The tools start, return data and AI to users.
This allows you to ask one question – if “where is this client connected?” – and accepting answers from multiple clusters and organizations.
IMPERATIVE Tools vs. Declarative tools on MCP servers
Demo explains two types of tools supported by MCP servers:
- Imperative They are predefined sequences written in Ansible, Terraform or Python. They are best suitable for registration, where the notch and strict order of implementation are important.
- Declarative tools They are automatically generated from Yaml files and are ideal for reading -intensive tasks such as inventory, event search or conformity control. They also support paging with offset and parameters.
Gabi shares examples of both types and demonstrates their use in real scenarios, such as firmware checks and discovery of customers between domains.
Troubleshooting and compliance with regulations using generative flows AI
In addition to calling with one tool, it supports MCP multi -stage workflows. These generative flows AI allow you to:
- Correlation of events
- Identify root causes such as BGP flaps
- Start compliance checks or collecting telemetry across websites
- APRY GUARDRAils for changes, securing only trusted Runbops for configuration actions
The MCP client learns from the tool formulas and can design new tools based on the Freque API call.
How to start and what will be next
This demo provides a clear and practical introduction to MCP for anyone working in Netops or Devops. Better understanding you get:
- Why does MCP today matter today
- How to connect MCP to your Cisco platforms
- Types of tools and workflows that support
- How to structure your own tools using Yaml or SDKS
Follow complete playback:
Subscribe to Cisco Devnet on YouTube to be informed about when the new demonstration will be released.
(Tagstotranslate) Cisco Catalyst (T) Cisco Meraki (T) MCP Server (T) network network automation