AWS re:Invent 2025 - AWS Cloud WAN MCP Server: Transform network operations with GenAI (NET331)

AWS re:Invent 2025 - AWS Cloud WAN MCP Server: Transform network operations with GenAI (NET331)

Overview

The presentation discussed the evolution of network management and operations, highlighting the shift towards a more collaborative approach between core networking and DevOps teams. The speakers introduced the AWS Networking MCP (Model Context Protocol) Server, an open-source tool designed to simplify network troubleshooting and analysis using AI-powered agents.

Challenges with Traditional Network Management

• Complex network architectures with multiple regions, VPCs, network function groups, and inspection VPCs
• Time-consuming and error-prone manual troubleshooting using the AWS Console, CLI, and Python scripts
• Difficulty in understanding the end-to-end network topology and connectivity

Introducing the AWS Networking MCP Server

• Designed to provide a comprehensive, AI-powered solution for network troubleshooting and analysis
• Composed of a set of specialized tools that can be orchestrated by an AI agent to achieve complex tasks
• Leverages the Model Context Protocol (MCP) to connect the AI agent to external data sources and tools

Key Design Principles

1. Modular Tool Design: The MCP server is built with discrete, purpose-driven tools rather than a monolithic script. This allows for greater flexibility and extensibility.
2. Agent-Driven Workflows: The AI agent is responsible for orchestrating the tool chain and decision-making, leveraging the capabilities of the language model.
3. Error Handling and Reporting: The tools provide detailed error handling and reporting, ensuring the agent and language model have the necessary information to make informed decisions.
4. Persona and Context: The use of an agent file with a defined persona and tool usage policy helps guide the agent's behavior and decision-making.

Demonstration: Finding an IP Address

• The presenters demonstrated how to use the AI agent and the MCP server to quickly create a custom tool for finding an IP address and its associated details.
• The tool was built using the agent's coding capabilities, leveraging the MCP server's tools and APIs.
• The resulting tool was able to search across multiple AWS regions to find the IP address and provide detailed information about the associated resources.

Demonstration: Path Tracing

• The presenters showcased the path tracing capabilities of the AWS Networking MCP Server, which can analyze the end-to-end network path between two IP addresses.
• The path tracing process involves calling multiple discrete tools within the MCP server, with the agent orchestrating the workflow and decision-making.
• The output is a detailed report in Markdown format, including mermaid diagrams that visually represent the network topology and connectivity.

Demonstration: Analyzing Cloud WAN Policies

• The presenters demonstrated how the MCP server can be used to analyze the configuration and policies of an AWS Cloud WAN network.
• The tool can provide detailed information about the network function groups, connectivity, and routing policies, with the output formatted in Markdown and including mermaid diagrams.

Key Takeaways

1. The AWS Networking MCP Server provides a powerful, AI-driven solution for network troubleshooting and analysis, addressing the challenges of complex network environments.
2. The modular tool design and agent-driven workflows enable greater flexibility, extensibility, and intelligent decision-making compared to traditional scripting approaches.
3. The use of an agent file with a defined persona and context helps guide the AI agent's behavior and ensures the desired outcomes are achieved.
4. The detailed reporting capabilities, including Markdown documentation and mermaid diagrams, greatly enhance the visibility and understanding of network configurations and connectivity.
5. The open-source nature of the AWS Networking MCP Server encourages community involvement and contributions, allowing for ongoing enhancements and the addition of support for new AWS network services.

Business Impact

• The AWS Networking MCP Server can significantly improve the efficiency and effectiveness of network operations, reducing the time and effort required for troubleshooting and analysis.
• The AI-powered capabilities can help bridge the gap between core networking and DevOps teams, fostering greater collaboration and knowledge sharing.
• The detailed reporting and visualization features can enhance communication and decision-making for network-related initiatives, improving overall network management and optimization.

Real-world Applications

• Troubleshooting complex network issues, such as connectivity problems, security control failures, and routing anomalies
• Analyzing the architecture and configuration of AWS Cloud WAN networks, including network function groups and connectivity policies
• Proactively monitoring network health and identifying potential issues before they impact business operations
• Onboarding and training new network engineers by providing comprehensive documentation and visualizations of the network environment