Metrics API

The Metrics API enables programmatic management of data quality metrics. Use it to track trends in your data over time, detect anomalies, and integrate quality monitoring into your pipelines.

Endpoints

Method	Endpoint	Description
GET	`/api/v1/sdk/metrics/{asset_id}/summary`	Get metrics summary for an asset
GET	`/api/v1/sdk/metrics/{asset_id}`	List metrics for an asset
GET	`/api/v1/sdk/metrics/{asset_id}/{metric_id}`	Get metric details
POST	`/api/v1/sdk/metrics/{asset_id}`	Create a new metric
PATCH	`/api/v1/sdk/metrics/{asset_id}/{metric_id}`	Update a metric
DELETE	`/api/v1/sdk/metrics/{asset_id}/{metric_id}`	Delete a metric
POST	`/api/v1/sdk/metrics/{asset_id}/{metric_id}/capture`	Trigger metric capture
GET	`/api/v1/sdk/metrics/{asset_id}/{metric_id}/snapshots`	List metric snapshots

Metric Types

Type	Description	Requires Column
`row_count`	Total row count of the table	No
`null_percent`	Percentage of null values	Yes
`distinct_count`	Count of distinct values	Yes
`duplicate_count`	Count of duplicate values	Yes
`min_value`	Minimum numeric value	Yes
`max_value`	Maximum numeric value	Yes
`mean`	Average numeric value	Yes
`percentile`	Percentile value (requires `percentile_value`)	Yes

Get Metrics Summary

GET /api/v1/sdk/metrics/{asset_id}/summary

Returns aggregate metrics statistics for an asset.

curl -H "Authorization: Bearer aa_live_xxx" \
  "https://api.anomalyarmor.ai/api/v1/sdk/metrics/550e8400-e29b-41d4-a716-446655440000/summary"

from anomalyarmor import Client

client = Client()
summary = client.metrics.summary("550e8400-e29b-41d4-a716-446655440000")
print(f"Active metrics: {summary.active_metrics}/{summary.total_metrics}")
print(f"Health: {summary.health_percentage}%")

armor metrics summary 550e8400-e29b-41d4-a716-446655440000

Response

{
  "data": {
    "total_metrics": 15,
    "active_metrics": 12,
    "total_checks": 8,
    "passing": 6,
    "failing": 1,
    "warning": 1,
    "error": 0,
    "health_percentage": 87.5
  }
}

List Metrics

GET /api/v1/sdk/metrics/{asset_id}

Query Parameters

Parameter	Type	Default	Description
`metric_type`	string	-	Filter by type (e.g., `row_count`, `null_percent`)
`is_active`	boolean	-	Filter by active status
`limit`	integer	50	Max results (max: 100)
`offset`	integer	0	Results to skip

curl -H "Authorization: Bearer aa_live_xxx" \
  "https://api.anomalyarmor.ai/api/v1/sdk/metrics/550e8400-e29b-41d4-a716-446655440000?metric_type=null_percent&limit=10"

from anomalyarmor import Client

client = Client()

# List all metrics
metrics = client.metrics.list("550e8400-e29b-41d4-a716-446655440000")

# Filter by type
null_metrics = client.metrics.list(
    "550e8400-e29b-41d4-a716-446655440000",
    metric_type="null_percent",
)
for m in null_metrics:
    print(f"{m.table_path}.{m.column_name}: {m.metric_type}")

armor metrics list 550e8400-e29b-41d4-a716-446655440000 --type null_percent

Response

{
  "data": {
    "items": [
      {
        "id": "m_550e8400-e29b-41d4-a716-446655440001",
        "internal_id": 123,
        "asset_id": 456,
        "table_path": "snowflake.prod.warehouse.orders",
        "column_name": "customer_email",
        "metric_type": "null_percent",
        "capture_interval": "daily",
        "sensitivity": 3,
        "is_active": true,
        "created_at": "2026-01-01T10:00:00Z"
      }
    ]
  },
  "pagination": {
    "total": 15,
    "limit": 50,
    "offset": 0,
    "has_more": false
  }
}

Get Metric Details

GET /api/v1/sdk/metrics/{asset_id}/{metric_id}

Query Parameters

Parameter	Type	Default	Description
`include_snapshots`	boolean	true	Include recent snapshots
`snapshot_limit`	integer	30	Max snapshots to include

curl -H "Authorization: Bearer aa_live_xxx" \
  "https://api.anomalyarmor.ai/api/v1/sdk/metrics/550e8400-e29b-41d4-a716-446655440000/m_550e8400-e29b-41d4-a716-446655440001?include_snapshots=true&snapshot_limit=10"

from anomalyarmor import Client

client = Client()
metric = client.metrics.get(
    "550e8400-e29b-41d4-a716-446655440000",
    "m_550e8400-e29b-41d4-a716-446655440001",
    include_snapshots=True,
    snapshot_limit=10,
)
print(f"Type: {metric.metric_type}")
print(f"Table: {metric.table_path}")
print(f"Active: {metric.is_active}")

armor metrics get 550e8400-e29b-41d4-a716-446655440000 m_550e8400-e29b-41d4-a716-446655440001

Response

{
  "data": {
    "id": "m_550e8400-e29b-41d4-a716-446655440001",
    "internal_id": 123,
    "asset_id": 456,
    "table_path": "snowflake.prod.warehouse.orders",
    "column_name": "customer_email",
    "metric_type": "null_percent",
    "capture_interval": "daily",
    "sensitivity": 3,
    "is_active": true,
    "created_at": "2026-01-01T10:00:00Z"
  }
}

Create Metric

POST /api/v1/sdk/metrics/{asset_id}

Requires read-write or admin scope.

Request Body

Field	Type	Required	Description
`metric_type`	string	Yes	Metric type (see table above)
`table_path`	string	Yes	Full table path (catalog.schema.table)
`column_name`	string	For column metrics	Column name
`capture_interval`	string	No	`hourly`, `daily`, `weekly` (default: `daily`)
`sensitivity`	float	No	Anomaly detection sensitivity (default: 1.0)
`group_by_columns`	array	No	Columns to group by
`percentile_value`	float	No	Percentile value (for `percentile` type)

curl -X POST -H "Authorization: Bearer aa_live_xxx" \
  -H "Content-Type: application/json" \
  -d '{
    "metric_type": "null_percent",
    "table_path": "snowflake.prod.warehouse.orders",
    "column_name": "customer_email",
    "capture_interval": "daily"
  }' \
  "https://api.anomalyarmor.ai/api/v1/sdk/metrics/550e8400-e29b-41d4-a716-446655440000"

from anomalyarmor import Client

client = Client()

# Create a null percentage metric
metric = client.metrics.create(
    "550e8400-e29b-41d4-a716-446655440000",
    metric_type="null_percent",
    table_path="snowflake.prod.warehouse.orders",
    column_name="customer_email",
    capture_interval="daily",
)
print(f"Created metric: {metric.id}")

# Create a row count metric
row_metric = client.metrics.create(
    "550e8400-e29b-41d4-a716-446655440000",
    metric_type="row_count",
    table_path="snowflake.prod.warehouse.orders",
)

armor metrics create 550e8400-e29b-41d4-a716-446655440000 \
  --type null_percent \
  --table snowflake.prod.warehouse.orders \
  --column customer_email

Response

{
  "data": {
    "id": "m_550e8400-e29b-41d4-a716-446655440001",
    "internal_id": 123,
    "asset_id": 456,
    "table_path": "snowflake.prod.warehouse.orders",
    "column_name": "customer_email",
    "metric_type": "null_percent",
    "capture_interval": "daily",
    "sensitivity": 1.0,
    "is_active": true,
    "created_at": "2026-01-04T10:30:00Z"
  }
}

Update Metric

PATCH /api/v1/sdk/metrics/{asset_id}/{metric_id}

Requires read-write or admin scope.

Request Body

Field	Type	Description
`is_active`	boolean	Whether metric is active
`capture_interval`	string	Capture interval
`sensitivity`	float	Anomaly detection sensitivity

curl -X PATCH -H "Authorization: Bearer aa_live_xxx" \
  -H "Content-Type: application/json" \
  -d '{"is_active": false}' \
  "https://api.anomalyarmor.ai/api/v1/sdk/metrics/550e8400-e29b-41d4-a716-446655440000/m_550e8400-e29b-41d4-a716-446655440001"

from anomalyarmor import Client

client = Client()
metric = client.metrics.update(
    "550e8400-e29b-41d4-a716-446655440000",
    "m_550e8400-e29b-41d4-a716-446655440001",
    is_active=False,
    sensitivity=2.0,
)

armor metrics update 550e8400-e29b-41d4-a716-446655440000 \
  m_550e8400-e29b-41d4-a716-446655440001 \
  --inactive

Delete Metric

DELETE /api/v1/sdk/metrics/{asset_id}/{metric_id}

Requires read-write or admin scope.

curl -X DELETE -H "Authorization: Bearer aa_live_xxx" \
  "https://api.anomalyarmor.ai/api/v1/sdk/metrics/550e8400-e29b-41d4-a716-446655440000/m_550e8400-e29b-41d4-a716-446655440001"

from anomalyarmor import Client

client = Client()
client.metrics.delete(
    "550e8400-e29b-41d4-a716-446655440000",
    "m_550e8400-e29b-41d4-a716-446655440001",
)

armor metrics delete 550e8400-e29b-41d4-a716-446655440000 \
  m_550e8400-e29b-41d4-a716-446655440001 --yes

Response

{
  "data": {
    "success": true,
    "message": "Metric deleted"
  }
}

Trigger Metric Capture

POST /api/v1/sdk/metrics/{asset_id}/{metric_id}/capture

Requires read-write or admin scope.

Triggers an immediate capture of the metric value.

curl -X POST -H "Authorization: Bearer aa_live_xxx" \
  "https://api.anomalyarmor.ai/api/v1/sdk/metrics/550e8400-e29b-41d4-a716-446655440000/m_550e8400-e29b-41d4-a716-446655440001/capture"

from anomalyarmor import Client

client = Client()
result = client.metrics.capture(
    "550e8400-e29b-41d4-a716-446655440000",
    "m_550e8400-e29b-41d4-a716-446655440001",
)
print(f"Captured {result.get('snapshot_count', 0)} snapshots")

armor metrics capture 550e8400-e29b-41d4-a716-446655440000 \
  m_550e8400-e29b-41d4-a716-446655440001

Response

{
  "data": {
    "snapshot_count": 1,
    "snapshots": [
      {
        "id": 789,
        "value": 2.5,
        "captured_at": "2026-01-04T10:35:00Z",
        "is_anomaly": false,
        "status": "PASS"
      }
    ]
  }
}

List Metric Snapshots

GET /api/v1/sdk/metrics/{asset_id}/{metric_id}/snapshots

Query Parameters

Parameter	Type	Default	Description
`limit`	integer	100	Max results
`offset`	integer	0	Results to skip

curl -H "Authorization: Bearer aa_live_xxx" \
  "https://api.anomalyarmor.ai/api/v1/sdk/metrics/550e8400-e29b-41d4-a716-446655440000/m_550e8400-e29b-41d4-a716-446655440001/snapshots?limit=30"

from anomalyarmor import Client

client = Client()
snapshots = client.metrics.snapshots(
    "550e8400-e29b-41d4-a716-446655440000",
    "m_550e8400-e29b-41d4-a716-446655440001",
    limit=30,
)
for s in snapshots:
    anomaly = "*" if s.is_anomaly else ""
    print(f"{s.captured_at}: {s.value:.2f} {anomaly}")

Response

{
  "data": {
    "items": [
      {
        "id": 789,
        "metric_definition_id": 123,
        "value": 2.5,
        "captured_at": "2026-01-04T10:35:00Z",
        "is_anomaly": false,
        "z_score": 0.3,
        "status": "PASS"
      },
      {
        "id": 788,
        "metric_definition_id": 123,
        "value": 15.2,
        "captured_at": "2026-01-03T10:35:00Z",
        "is_anomaly": true,
        "z_score": 4.2,
        "status": "FAIL"
      }
    ]
  },
  "pagination": {
    "total": 90,
    "limit": 30,
    "offset": 0,
    "has_more": true
  }
}

Use Case: Monitor Row Count Trends

Track daily row counts to detect unexpected data volume changes:

from anomalyarmor import Client

client = Client()
asset_id = "550e8400-e29b-41d4-a716-446655440000"

# Create row count metric
metric = client.metrics.create(
    asset_id,
    metric_type="row_count",
    table_path="snowflake.prod.warehouse.orders",
    capture_interval="daily",
    sensitivity=2.0,  # Alert on 2+ standard deviations
)

# Trigger initial capture
result = client.metrics.capture(asset_id, metric.id)
print(f"Initial row count: {result['snapshots'][0]['value']}")

# Later: check for anomalies
snapshots = client.metrics.snapshots(asset_id, metric.id, limit=7)
anomalies = [s for s in snapshots if s.is_anomaly]
if anomalies:
    print(f"Found {len(anomalies)} anomalies in the last 7 captures")

Error Responses

Metric Not Found (404)

{
  "error": {
    "code": "METRIC_NOT_FOUND",
    "message": "Metric not found",
    "details": {"metric_id": "m_invalid-uuid"}
  }
}

Validation Error (400)

{
  "error": {
    "code": "VALIDATION_ERROR",
    "message": "column_name is required for null_percent metrics",
    "details": {"field": "column_name", "metric_type": "null_percent"}
  }
}

Forbidden (403)

{
  "error": {
    "code": "FORBIDDEN",
    "message": "Insufficient permissions. Required scope: read-write",
    "details": {"current_scope": "read-only", "required_scope": "read-write"}
  }
}

Common Questions

What’s the difference between a metric and a validity rule?

Metrics track numeric values over time (row counts, null percentages, mean, percentile) and alert on statistical anomalies via the sensitivity parameter. Validity rules (Validity API) enforce deterministic pass/fail constraints like NOT NULL, REGEX, or RANGE. Use metrics to catch drift, validity to catch explicit contract violations.

How does the sensitivity parameter affect anomaly detection?

Sensitivity is the z-score threshold for flagging a snapshot as anomalous, defaulting to 1.0. Raise it (e.g. 2.0 or 3.0) to reduce false positives on noisy data, lower it to catch subtler shifts. Each captured snapshot returns z_score and is_anomaly so you can tune in production.

Can I capture a metric on demand outside its scheduled interval?

Yes. POST /api/v1/sdk/metrics/{asset_id}/{metric_id}/capture (or client.metrics.capture(...)) triggers an immediate capture and returns the new snapshot with anomaly status. This is useful for backfilling after creating a metric and for post-deploy validation without waiting for the next scheduled run.

Which metric types require a column_name?

Column-level types (null_percent, distinct_count, duplicate_count, min_value, max_value, mean, percentile) require column_name. row_count operates on the whole table and ignores column_name. percentile additionally needs percentile_value in the request body.

API Overview

API Reference

API Guides

SDKs

CLI

Integrations

AI Agents

Endpoints

Metric Types

Get Metrics Summary

Response

List Metrics

Query Parameters

Response

Get Metric Details

Query Parameters

Response

Create Metric

Request Body

Response

Update Metric

Request Body

Delete Metric

Response

Trigger Metric Capture

Response

List Metric Snapshots

Query Parameters

Response

Use Case: Monitor Row Count Trends

Error Responses

Metric Not Found (404)

Validation Error (400)

Forbidden (403)

Common Questions

What’s the difference between a metric and a validity rule?

How does the sensitivity parameter affect anomaly detection?

Can I capture a metric on demand outside its scheduled interval?

Which metric types require a column_name?

​Endpoints

​Metric Types

​Get Metrics Summary

​Response

​List Metrics

​Query Parameters

​Response

​Get Metric Details

​Query Parameters

​Response

​Create Metric

​Request Body

​Response

​Update Metric

​Request Body

​Delete Metric

​Response

​Trigger Metric Capture

​Response

​List Metric Snapshots

​Query Parameters

​Response

​Use Case: Monitor Row Count Trends

​Error Responses

​Metric Not Found (404)

​Validation Error (400)

​Forbidden (403)

​Common Questions

​What’s the difference between a metric and a validity rule?

​How does the sensitivity parameter affect anomaly detection?

​Can I capture a metric on demand outside its scheduled interval?

​Which metric types require a column_name?

Endpoints

Metric Types

Get Metrics Summary

Response

List Metrics

Query Parameters

Response

Get Metric Details

Query Parameters

Response

Create Metric

Request Body

Response

Update Metric

Request Body

Delete Metric

Response

Trigger Metric Capture

Response

List Metric Snapshots

Query Parameters

Response

Use Case: Monitor Row Count Trends

Error Responses

Metric Not Found (404)

Validation Error (400)

Forbidden (403)

Common Questions

What’s the difference between a metric and a validity rule?

How does the sensitivity parameter affect anomaly detection?

Can I capture a metric on demand outside its scheduled interval?

Which metric types require a column_name?