Amazon Lookout for Metrics - Anomaly Detection Operations

Workflow capability for data science and operations teams to manage anomaly detectors, monitor metric anomalies, configure alerts, and provide detection feedback using Amazon Lookout for Metrics.

Run with Naftiko AmazonAnomaly DetectionMachine LearningMetricsMonitoringOperations

What You Can Do

POST
Create detector — Create a new anomaly detector
/v1/detectors
GET
List detectors — List all anomaly detectors
/v1/detectors
GET
Get detector — Get detector details
/v1/detectors/{id}
PUT
Update detector — Update detector configuration
/v1/detectors/{id}
DELETE
Delete detector — Delete a detector
/v1/detectors/{id}
GET
List anomalies — List anomaly group summaries
/v1/anomalies
GET
Get anomaly — Get anomaly group details
/v1/anomalies/{id}
POST
Create alert — Create an alert
/v1/alerts
GET
List alerts — List all alerts
/v1/alerts
POST
Submit feedback — Submit anomaly feedback
/v1/feedback
GET
Get feedback — Get anomaly feedback
/v1/feedback

MCP Tools

create-anomaly-detector

Create a new anomaly detector for a set of business metrics

list-anomaly-detectors

List all configured anomaly detectors

read-only idempotent
describe-anomaly-detector

Get configuration and status details of an anomaly detector

read-only idempotent
activate-anomaly-detector

Activate an anomaly detector to begin monitoring metrics

idempotent
deactivate-anomaly-detector

Stop an anomaly detector from monitoring metrics

idempotent
list-anomaly-group-summaries

List summaries of detected anomaly groups for investigation

read-only idempotent
get-anomaly-group

Get full details of a specific anomaly group including contributing metrics

read-only idempotent
list-anomaly-group-related-metrics

List all metrics that contributed to an anomaly group

read-only idempotent
create-alert

Create an alert to receive notifications when anomalies are detected

list-alerts

List all configured anomaly alerts

read-only idempotent
put-feedback

Submit feedback on anomaly detections to improve ML model accuracy

get-feedback

Retrieve previously submitted anomaly detection feedback

read-only idempotent

APIs Used

lookout-for-metrics