Atlas now manages Snowflake data metric functions with the new data_metric_function block.

The new data_metric_function block declares a Snowflake data metric function in a schema. Each arg is a table argument whose column blocks declare its columns.

Data Metric Functions

Like other functions, data metric functions are managed by the funcs mode, which is on by default.
A function takes the required schema and as attributes and arg blocks, the optional return, secure, comment, and depends_on attributes, and tag blocks. return can also be a block with type and null.
With permissions = true, permission blocks grant privileges on data metric functions.
With tags = true, tag blocks set tags on data metric functions.

Example

The env manages functions with the funcs mode, which is on by default, grants with permissions = true, and tags with tags = true:

atlas.hcl

env "local" {  url = "snowflake://user:password@account_identifier/DEMO_DB?warehouse=DEMO_WH"  dev = "snowflake://user:password@account_identifier/DEMO_DEV_DB?warehouse=DEMO_WH"  schema {    src = "file://schema.hcl"    mode "snowflake" {      funcs       = true      permissions = true      tags        = true    }  }}

The desired schema declares a secure data metric function that counts the rows where three columns are positive, tags it, and grants USAGE on it to the DQ_ANALYST database role, which exists outside Atlas:

schema.hcl

schema "PUBLIC" {}
tag "DATA_QUALITY_TIER" {  schema         = schema.PUBLIC  allowed_values = ["critical", "high", "low"]  comment        = "How urgent a failing data quality check is"}
data_metric_function "COUNT_POSITIVE_NUMBERS" {  schema = schema.PUBLIC  arg "ARG_T" {    column "ARG_C1" {      type = NUMBER    }    column "ARG_C2" {      type = NUMBER    }    column "ARG_C3" {      type = NUMBER    }  }  as      = <<-SQL  SELECT    COUNT(*)  FROM ARG_T  WHERE    ARG_C1 > 0    AND ARG_C2 > 0    AND ARG_C3 > 0  SQL  secure  = true  comment = "count positive numbers"  tag {    ref   = tag.DATA_QUALITY_TIER    value = "high"  }}
# DQ_ANALYST is a database role created outside Atlas: ref() only grants to it.permission {  to         = ref("database_role.DQ_ANALYST")  for        = data_metric_function.COUNT_POSITIVE_NUMBERS  privileges = [USAGE]}

Run the diff:

atlas migrate diff --env local

Atlas writes the migration:

migrations/20261111154802.sql

-- Create tag "DATA_QUALITY_TIER"CREATE TAG "PUBLIC"."DATA_QUALITY_TIER" ALLOWED_VALUES 'critical', 'high', 'low' COMMENT = 'How urgent a failing data quality check is';-- Create data metric function "COUNT_POSITIVE_NUMBERS"CREATE SECURE DATA METRIC FUNCTION "PUBLIC"."COUNT_POSITIVE_NUMBERS" ("ARG_T" TABLE ("ARG_C1" NUMBER, "ARG_C2" NUMBER, "ARG_C3" NUMBER)) RETURNS NUMBER COMMENT = 'count positive numbers' AS $$SELECT  COUNT(*)FROM ARG_TWHERE  ARG_C1 > 0  AND ARG_C2 > 0  AND ARG_C3 > 0$$;-- Modify data metric function "COUNT_POSITIVE_NUMBERS" to set tagsALTER FUNCTION "PUBLIC"."COUNT_POSITIVE_NUMBERS" (TABLE(NUMBER, NUMBER, NUMBER)) SET TAG PUBLIC.DATA_QUALITY_TIER = 'high';-- Grant on data metric function "COUNT_POSITIVE_NUMBERS" to "DQ_ANALYST"GRANT USAGE ON FUNCTION "PUBLIC"."COUNT_POSITIVE_NUMBERS" (TABLE(NUMBER, NUMBER, NUMBER)) TO DATABASE ROLE "DQ_ANALYST";

Getting Started

Snowflake support is part of Atlas Pro. Run atlas login to get started.

See the Snowflake HCL reference for every attribute and block.

Snowflake: Data Metric Functions

Data Metric Functions

Example

Getting Started