AWS CloudWatch MCP Server

About

## AWS CloudWatch MCP Server: Monitoring and Observability The **AWS CloudWatch MCP Server** integrates Amazon CloudWatch into Google Antigravity, enabling real-time monitoring, log analysis, and alerting for AWS resources directly from your development environment. ### Why AWS CloudWatch MCP? - **Unified Monitoring**: Access metrics, logs, and alarms across all AWS services in one place - **Real-Time Insights**: Query logs and metrics with sub-second latency for debugging - **Cost Visibility**: Monitor resource utilization to optimize AWS spending - **Automated Alerts**: Create and manage alarms for proactive incident response - **Dashboard Integration**: Build and view custom dashboards programmatically ### Key Features #### 1. Log Queries ```python # Query CloudWatch Logs with Insights results = await mcp.query_logs( log_group="/aws/lambda/my-function", query=""" fields @timestamp, @message | filter @message like /ERROR/ | sort @timestamp desc | limit 50 """, start_time="-1h" ) for log in results: print(f"[{log['timestamp']}] {log['message']}") ``` #### 2. Metrics Retrieval ```python # Get EC2 CPU utilization metrics = await mcp.get_metric_data( namespace="AWS/EC2", metric_name="CPUUtilization", dimensions={"InstanceId": "i-1234567890abcdef0"}, period=300, start_time="-6h", statistic="Average" ) for point in metrics["datapoints"]: print(f"{point['timestamp']}: {point['value']}%") ``` #### 3. Alarm Management ```python # Create a metric alarm await mcp.create_alarm( alarm_name="high-cpu-production", metric_name="CPUUtilization", namespace="AWS/EC2", threshold=80, comparison="GreaterThanThreshold", evaluation_periods=3, period=300, actions=["arn:aws:sns:us-east-1:123456789:alerts"] ) # List alarms in ALARM state alarms = await mcp.describe_alarms(state="ALARM") for alarm in alarms: print(f"ALARM: {alarm['name']} - {alarm['reason']}") ``` #### 4. Log Streaming ```python # Stream logs in real-time async for event in mcp.stream_logs( log_group="/aws/ecs/my-service", filter_pattern="ERROR" ): print(f"[{event['timestamp']}] {event['message']}") ``` ### Configuration ```json { "mcpServers": { "aws-cloudwatch": { "command": "npx", "args": ["-y", "@anthropic/mcp-cloudwatch"], "env": { "AWS_ACCESS_KEY_ID": "your-access-key", "AWS_SECRET_ACCESS_KEY": "your-secret-key", "AWS_REGION": "us-east-1" } } } } ``` ### Use Cases **Debug Production Issues**: Query application logs and correlate with metrics to quickly identify root causes of production incidents. **Performance Monitoring**: Track API latency, error rates, and throughput metrics across microservices to ensure SLA compliance. **Cost Optimization**: Monitor resource utilization patterns to right-size EC2 instances and optimize reserved capacity. **Proactive Alerting**: Create intelligent alarms that notify teams before issues impact users based on anomaly detection. The AWS CloudWatch MCP enables comprehensive AWS monitoring directly within your development environment.

Installation

{
  "mcpServers": {
    "aws-cloudwatch": {
      "mcpServers": {
        "awslabs.cloudwatch-mcp-server": {
          "env": {
            "AWS_PROFILE": "default",
            "FASTMCP_LOG_LEVEL": "ERROR"
          },
          "args": [
            "awslabs.cloudwatch-mcp-server@latest"
          ],
          "command": "uvx"
        }
      }
    }
  }
}

How to Use

1. 11. Set AWS_PROFILE to your AWS profile name
2. 22. Ensure you have valid AWS credentials configured
3. 33. Provides tools for alarm troubleshooting, metrics analysis, and log investigation