Official AWS integration — DynamoDB, S3, Lambda, EC2, and the full AWS cloud via natural language
The official AWS MCP Suite gives AI agents structured, permission-aware access to AWS services. Instead of juggling SDK boilerplate and IAM policies manually, your agent can query and manage infrastructure through natural language. Included servers: - DynamoDB: Query tables, get items, run PartiQL, manage streams - S3: List buckets, upload/download objects, manage policies and lifecycle rules - Lambda: Invoke functions, view logs, manage environment variables, deploy packages - EC2: Describe instances, start/stop/terminate, manage security groups - RDS/Aurora: Query databases, manage snapshots, monitor performance - CloudWatch: Query logs, set alarms, view metrics dashboards - IAM: Manage roles, policies, and access analysis The suite uses your existing AWS credentials (IAM roles, profiles, or environment variables) — no additional auth setup required. Scoped permissions ensure your agent only accesses what it needs.
from agents import Agent
from agents.mcp import MCPServerStdio
mcp_server = MCPServerStdio(
command="npx",
args=["-y", "@aws/mcp"],
env={
"AWS_ACCESS_KEY_ID": "AKIA...",
"AWS_SECRET_ACCESS_KEY": "your_secret",
"AWS_REGION": "us-east-1",
},
)
agent = Agent(
name="My Agent",
model="gpt-4o",
mcp_servers=[mcp_server],
)If you maintain AWS MCP Suite, add this badge to your README to show it's verified on CuratedMCP:
[](https://curatedmcp.com/marketplace/aws-mcp-suite)