AWS MCP Server with Agno Agent

What is AWS MCP Server with Agno Agent?

It is an MCP (Model Context Protocol) server that integrates AWS tools (listing security groups, S3 buckets, analyzing VPC connections) using FastMCP and the Agno framework. It enables remote clients to interact with these AWS operations via standardized MCP endpoints and includes an interactive agent that uses local Ollama models.

How to use AWS MCP Server with Agno Agent?

After installing Python 3.11+, configuring AWS credentials, and installing Ollama, clone the repository and install dependencies. Run the MCP server with python src/aws-security-mcp-server.py. Then run the interactive agent with python src/aws-demo-agent.py. Alternatively, connect any MCP-compliant client (e.g., using Agno’s MCPTools) to the SSE endpoint at http://127.0.0.1:8000/sse/.

Key features of AWS MCP Server with Agno Agent

• AWS Security Group Tool: Lists security groups with inbound/outbound rules.
• AWS S3 Bucket Tool: Lists S3 buckets with region, creation date, and access policy.
• AWS VPC Connections Tool: Analyzes VPC peering, endpoints, transit gateways, route tables, and network ACLs.
• MCP Server with SSE transport: Real‑time communication for MCP‑compliant clients.
• Interactive Agno Agent: Task management, command history, and real‑time tool call streaming.
• Ollama Integration: Uses local LLMs, reducing dependency on external APIs.

Use cases of AWS MCP Server with Agno Agent

• Query AWS security groups across regions from any MCP client.
• List and inspect S3 bucket properties and policies.
• Analyze VPC network connections including peering, endpoints, and route tables.
• Build custom AI agents that can manage AWS infrastructure.
• Run parallel AWS queries in a terminal with real‑time streaming.

FAQ from AWS MCP Server with Agno Agent

What dependencies are required?

Python 3.11+, AWS credentials configured for boto3, Ollama installed locally, and Python packages: boto3, agno, fastmcp, rich, click.

How does the server communicate with clients?

It uses FastMCP with Server‑Sent Events (SSE) transport, listening on 127.0.0.1:5678 (the SSE URL is typically http://127.0.0.1:8000/sse/).

Can I use external AI models instead of Ollama?

The project demonstrates only local Ollama integration. Using other models would require modifying the agent code.

What AWS actions can the server perform?

It can list security groups, list S3 buckets, and analyze VPC connections (peering, endpoints, transit gateways, route tables, network ACLs). All operations use boto3 with the configured AWS credentials.

Is this project open source and free?

Yes, it is licensed under the MIT License (described as “Free for all!”).