Connect AWS to AI: Let Claude and Cursor Access the AWS Console, CloudWatch, and S3

Why Connect AWS to AI?

Your AWS Console is where infrastructure lives. CloudWatch logs, S3 buckets, Lambda functions, EC2 instances, RDS databases, IAM roles. When you're debugging in your editor and need to check a log stream, verify an S3 object, or see if a Lambda is erroring, you open the Console, navigate through nested menus, and manually look things up.

With an MCP server pointed at your AWS Console, AI agents can query your infrastructure directly. Check logs, list objects, monitor services without switching to the browser.

What AI Agents Can Do With AWS Access

Once connected, Claude, Cursor, or Codex can:

• Check CloudWatch logs — "What errors are in the payment-service log group from the last hour?"
• Browse S3 — "What files are in the exports bucket under /reports/2026-06/?"
• Monitor Lambda — "What's the error rate for the processOrder function today?"
• Check EC2 — "What instances are running in us-east-1 and what's their status?"
• Review IAM — "What policies are attached to the deployment role?"
• Inspect RDS — "What's the current connection count on the production database?"

How It Works

1. Open DataFaucet and enter your AWS Console URL
2. Browse the Console — click through CloudWatch, S3, Lambda, EC2, whatever you need
3. DataFaucet captures the read operations as MCP tools
4. Deploy — your AI agents now have structured access to query AWS

You're browsing the AWS Console web UI, not sharing IAM credentials or access keys. DataFaucet captures HTTP interactions and replays them. Your AWS credentials stay in your browser session.

Example Queries

"What's in the CloudWatch logs for /aws/lambda/processPayment in the last 30 minutes?"
"List all objects in s3://my-bucket/data/2026-06-01/"
"What Lambda functions had errors today?"
"What EC2 instances are running and what's their CPU utilization?"
"Show me the most recent CloudFormation stack events"

When This Saves Time

• Incident response — pull CloudWatch logs alongside your code changes
• Deployment verification — check Lambda status and S3 artifacts after deploy
• Infrastructure debugging — ask about resource state without navigating Console
• Cost investigation — quick checks on instance types and utilization

Security Model

DataFaucet connects through the AWS Console web UI. Access is scoped to whatever your IAM user or role can see in the Console. If your permissions only allow read access to certain services, the MCP server has the same constraints. No access keys or secrets are stored separately.

Getting Started

Point DataFaucet at your AWS Console URL (console.aws.amazon.com). Browse through the services you want AI to access — CloudWatch, S3, Lambda, EC2. Deploy. Your agents can now query AWS infrastructure from your terminal.

Create your AWS MCP server →