Fabian
12/16/2025, 8:17 AMkapa.ai
12/16/2025, 8:17 AMkapa.ai
12/16/2025, 8:18 AMelementary schema (or whatever schema you configured for the package). This includes models like elementary_test_results, dbt_models, dbt_sources, dbt_tests, dbt_source_freshness_results, etc. [[Elementary dbt package](https://docs.elementary-data.com/data-tests/dbt/dbt-package); [Elementary in production](https://docs.elementary-data.com/oss/deployment-and-configuration/elementary-in-production)] Report “monitored tables,” for example, are built by unioning dbt_models and dbt_sources and joining to dbt_tests. [[Slack: monitored tables](https://elementary-community.slack.com/archives/C02CTC89LAX/p1707205953561139)] 2. dbt artifacts documentation in the repo Support recommends using the models/dbt_artifacts.yml file in the open‑source repo as the best reference for the structure and purpose of these tables. [[Slack: schema request](https://elementary-community.slack.com/archives/C02CTC89LAX/p1707205953561139)] You can find it here: <https://github.com/elementary-data/dbt-data-reliability/blob/master/models/dbt_artifacts.yml> 3. What types of data are stored At a high level Elementary stores, in that schema: metadata (models, sources, columns), configurations, run results, query logs, table metrics, aggregated metrics, and (optionally) samples of failed rows. [[Security & privacy – what information is collected](https://docs.elementary-data.com/cloud/general/security-and-privacy#what-information-does-elementary-collect)] 4. No official full ERD / foreign‑key map Multiple community questions asked for a complete data model / schema diagram, and the official guidance was to rely on the dbt‑package documentation and inspect the warehouse tables directly; a full relational diagram is not provided in the public docs. [[Slack: data model request](https://elementary-community.slack.com/archives/C02CTC89LAX/p1689961625483339); [Slack: schema for dashboards](https://elementary-community.slack.com/archives/C02CTC89LAX/p1690187547399799)] So: there is no single published “elementary data model” document, but you can reconstruct it by: - Exploring all tables in your configured Elementary schema in the warehouse, and - Using dbt_artifacts.yml in the repo as the authoritative description of each model and its fields.