Check Types¶

Provero includes 16 built-in check types organized by category. Each check can be written in shorthand or expanded YAML form.

Completeness¶

not_null¶

Verifies that column(s) contain no null values.

Parameter	Type	Required	Description
column(s)	string or list	Yes	Column name or list of column names

Default severity: critical

ShorthandExpanded

checks:
  - not_null: order_id
  - not_null: [order_id, customer_id, amount]

checks:
  - not_null:
      column: order_id
      severity: warning

completeness¶

Checks that a column meets a minimum percentage of non-null values.

Parameter	Type	Required	Default	Description
`column`	string	Yes		Column to check
`min`	number	No	`0.95`	Minimum completeness ratio (0 to 1)

Default severity: critical

checks:
  - completeness:
      column: email
      min: 0.90

Uniqueness¶

unique¶

Verifies that a column has no duplicate values.

Parameter	Type	Required	Description
column	string	Yes	Column name

Default severity: critical

checks:
  - unique: order_id

unique_combination¶

Checks that a combination of columns is unique (composite key).

Parameter	Type	Required	Description
columns	list	Yes	List of column names (minimum 2)

Default severity: critical

checks:
  - unique_combination: [date, store_id]

Validity¶

accepted_values¶

Ensures column values are within an allowed set.

Parameter	Type	Required	Description
`column`	string	Yes	Column to check
`values`	list	Yes	Allowed values (strings, numbers, or booleans)

Default severity: critical

checks:
  - accepted_values:
      column: status
      values: [pending, shipped, delivered, cancelled]

range¶

Verifies that numeric values fall within min/max bounds.

Parameter	Type	Required	Description
`column`	string	Yes	Column to check
`min`	number	No	Minimum allowed value
`max`	number	No	Maximum allowed value

At least one of min or max should be specified.

Default severity: critical

checks:
  - range:
      column: amount
      min: 0
      max: 100000

regex¶

Validates that column values match a regular expression pattern.

Parameter	Type	Required	Description
`column`	string	Yes	Column to check
`pattern`	string	Yes	Regular expression pattern

Works across databases: uses regexp_matches() on DuckDB, ~ on PostgreSQL, and REGEXP on MySQL/SQLite.

Default severity: warning

checks:
  - regex:
      column: email
      pattern: "^[^@]+@[^@]+\\.[^@]+$"

email_validation¶

Validates that column values are valid email addresses using a standard email regex pattern.

Parameter	Type	Required	Description
`column`	string	Yes	Column to check

Works across databases: uses regexp_matches() on DuckDB, ~ on PostgreSQL, and REGEXP on MySQL/SQLite.

Default severity: warning

checks:
  - email_validation:
      column: email

type¶

Checks that a column's data type matches the expected type.

Parameter	Type	Required	Description
`column`	string	Yes	Column to check
`expected`	string	Yes	Expected type: `integer`, `float`, `string`, `boolean`, `date`, `timestamp`

Type names are normalized across databases. For example, int, int4, bigint, and smallint all match integer.

Default severity: critical

checks:
  - type:
      column: amount
      expected: float

Freshness¶

freshness¶

Checks that the most recent row is within a time threshold.

Parameter	Type	Required	Default	Description
`column`	string	Yes		Timestamp column
`max_age`	string	Yes	`24h`	Maximum age. Format: `30m`, `24h`, `7d`

Default severity: critical

checks:
  - freshness:
      column: updated_at
      max_age: 24h

latency¶

Measures the time difference between two timestamp columns. Useful for detecting pipeline delays.

Parameter	Type	Required	Default	Description
`source_column`	string	Yes		Start timestamp (e.g., event time)
`target_column`	string	Yes		End timestamp (e.g., load time)
`max_latency`	string	No	`1h`	Maximum acceptable latency

Default severity: warning

checks:
  - latency:
      source_column: created_at
      target_column: processed_at
      max_latency: 1h

Volume¶

row_count¶

Validates that the table row count is within expected bounds.

Parameter	Type	Required	Default	Description
`min`	integer	No	`0`	Minimum row count
`max`	integer	No		Maximum row count

Default severity: critical

checks:
  - row_count:
      min: 1
      max: 1000000

row_count_change¶

Compares the current row count against the previous run. Requires the result store to be enabled (default).

Parameter	Type	Required	Default	Description
`max_decrease`	string	No	`50%`	Maximum allowed decrease percentage
`max_increase`	string	No	`500%`	Maximum allowed increase percentage

Default severity: warning

checks:
  - row_count_change:
      max_decrease: "10%"
      max_increase: "200%"

Anomaly Detection¶

anomaly¶

Statistical anomaly detection on historical metrics. Compares the current value against stored history using one of three methods.

Parameter	Type	Required	Default	Description
`metric`	string	Yes		Metric to check: `row_count`, `null_count`, `null_rate`, `distinct_count`, `mean`, `min`, `max`
`column`	string	No		Column for column-level metrics
`method`	string	No	`mad`	Detection method: `zscore`, `mad`, `iqr`
`sensitivity`	string	No	`medium`	Sensitivity level
`threshold`	float	No		Direct threshold override

Default severity: warning

checks:
  - anomaly:
      column: daily_revenue
      metric: mean
      method: mad
      sensitivity: medium

Detection methods:

Method	Description	Best for
`zscore`	Standard Z-Score	Normally distributed metrics
`mad`	Median Absolute Deviation	Robust to outliers
`iqr`	Interquartile Range	Skewed distributions

All methods are implemented using Python stdlib only (no scipy dependency). Anomaly detection uses the result store to compare current values against historical data. Run provero run regularly to build up the baseline.

Referential Integrity¶

referential_integrity¶

Validates that all non-null values in a column exist in a reference table's column (foreign key validation). Orphaned rows where the FK value does not exist in the referenced table cause the check to fail. NULL values are excluded, as they represent optional relationships.

Parameter	Type	Required	Description
`column`	string	Yes	FK column in the source table
`reference_table`	string	Yes	Referenced table name
`reference_column`	string	Yes	Referenced column (usually the primary key)

Default severity: critical

checks:
  - referential_integrity:
      column: customer_id
      reference_table: customers
      reference_column: id

Custom¶

custom_sql¶

Runs a custom SQL query that must return a truthy value to pass.

Parameter	Type	Required	Description
`query`	string	Yes	SQL query returning a single value
`name`	string	No	Custom name for the check

Default severity: critical

ShorthandExpanded

checks:
  - custom_sql: "SELECT COUNT(*) > 0 FROM orders"

checks:
  - custom_sql:
      name: positive_revenue
      query: "SELECT SUM(amount) > 0 FROM orders"