03-aws-assistant-mcp

🚀 AWS Assistant

Overview

AWS Assistant is a sophisticated multi-agent system designed to provide comprehensive insights and assistance for AWS-related queries. Leveraging advanced AI technologies and model context protocol (MCP), this assistant offers three primary specialized agents:

Feature	Description
Agent Structure	Multi-agent architecture
Native Tools	think, file_write, python_repl, shell
Custom Agents	aws_documentation_researcher, graph_creater, aws_cost_assistant
MCP Servers	AWS Cost Explorer, AWS Documentation
Model Provider	Amazon Bedrock

Caution

python_repl and shell tools can run commands in your environment. Make sure to run this sample in a sandbox environment.

Key Features

1. AWS Documentation Researcher 📚

• Searches and analyzes AWS documentation
• Provides detailed, source-cited explanations
• Ideal for technical and procedural AWS queries

2. AWS Cost Assistant 💰

• Analyzes AWS account spending
• Generates detailed cost breakdowns
• Supports queries about service expenditures, regional costs, and usage patterns

3. Graph Creator 📊

• Visualizes complex AWS cost and usage data
• Generates interactive graphs using Plotly
• Transforms raw data into meaningful visual representations

Getting Started

1. Install uv.

2. Install Docker and make sure the Docker daemon is running. Checkout Docker Desktop, and explore Docker Desktop.

3. Setup aws-cost-explorer-mcp-server MCP server.

4. Set up AWS credentials in .env using .env.example.

5. Complete prerequisites for aws-cost-explorer-mcp-server.

   [!DISCLAIMER] python_repl tool uses plotly to create graphs. Make sure to pip install plotly before using the Graph Creater Agent

6. Run the AWS Assistant using uv run main.py

Example Queries

• "Explain AWS Lambda triggers"
• "What's my AWS spending this month?"
• "Create a graph of my service costs"