can dbt_artifacts be kicked off on-run-start as well ?

[yxl1108]

dbt_artifacts upload_results writes the run_started_at into the logging table. I am thinking of using that time for Snowflake time travel, so in the dbt project run, all source tables will be queried at a consistent time.

example in a model file would be

{% set dbt_run_time = get_dbt_run_time() %}

select 
    *
from {{ source("schema", "table") }}
    at(timestamp => '{{ dbt_run_time }}'::timestamp) as a

where get_dbt_run_time() is a macro to get the latest run_started_at from invocations table for the project_name, provided the invocation entry is created before any model runs

[yxl1108]

figured out how to do this, run_started_at is a jinja variable available in dbt.