Kind Tags#

Kind tags can help you quickly identify the underlying system or technology used for a given asset in the Dagster UI. These tags allow you to organize assets within a Dagster project and can be used to filter or search across all your assets.

Adding kinds to an asset#

You may add up to three kinds to the kinds argument, which can be useful to represent multiple technologies or systems that an asset is associated with. For example, an asset which is built by Python code and stored in Snowflake can be tagged with both python and snowflake kinds.


from dagster import asset


@asset(kinds={"python", "snowflake"})
def my_asset():
    pass

Kinds can also be specified on an AssetSpec, for use in multi-assets:


from dagster import AssetSpec, multi_asset


@multi_asset(
    specs=[
        AssetSpec("foo", kinds={"python", "snowflake"}),
        AssetSpec("bar", kinds={"python", "postgres"}),
    ]
)
def my_multi_asset():
    pass

Under the hood, these kind inputs are stored as tags on the asset. For more information, see Tags.

When viewing the asset in the lineage view, the attached kinds will be visible at the bottom the asset.

Adding compute kinds to assets#

Using compute_kind has been superseded by the kinds argument. We recommend using the kinds argument instead.

You can add a single compute kind to any asset by providing a value to the compute_kind argument.


@asset(compute_kind="dbt")
def my_asset():
    pass

Supported Icons#

Some kinds are given a branded icon in the UI. We currently support nearly 200 unique technology icons.

Value	Icon
`airbyte`
`airflow`
`airliftmapped`
`airtable`
`athena`
`atlan`
`aws`
`awsstepfunction`
`awsstepfunctions`
`axioma`
`azure`
`azureml`
`bigquery`
`cassandra`
`catboost`
`celery`
`census`
`chalk`
`claude`
`clickhouse`
`cockroachdb`
`collibra`
`cplus`
`cplusplus`
`csharp`
`cube`
`dask`
`databricks`
`datadog`
`datahub`
`dbt`
`deltalake`
`denodo`
`discord`
`dlt`
`dlthub`
`docker`
`doris`
`druid`
`duckdb`
`elasticsearch`
`excel`
`facebook`
`fivetran`
`flink`
`gcp`
`gcs`
`gemini`
`github`
`gitlab`
`go`
`google`
`googlecloud`
`googledrive`
`googlesheets`
`graphql`
`greatexpectations`
`hackernews`
`hackernewsapi`
`hadoop`
`hashicorp`
`hex`
`hightouch`
`hudi`
`huggingface`
`huggingfaceapi`
`iceberg`
`icechunk`
`impala`
`instagram`
`ipynb`
`java`
`javascript`
`jupyter`
`k8s`
`kafka`
`kubernetes`
`lakefs`
`lightgbm`
`linear`
`linkedin`
`llama`
`looker`
`mariadb`
`matplotlib`
`meltano`
`meta`
`metabase`
`microsoft`
`minio`
`minstral`
`mlflow`
`modal`
`mongodb`
`montecarlo`
`mysql`
`net`
`noteable`
`notion`
`numpy`
`omni`
`openai`
`openmetadata`
`optuna`
`oracle`
`pagerduty`
`pandas`
`pandera`
`papermill`
`papertrail`
`parquet`
`pinot`
`plotly`
`plural`
`polars`
`postgres`
`postgresql`
`powerbi`
`prefect`
`presto`
`pulsar`
`pyspark`
`python`
`pytorch`
`pytorchlightning`
`r`
`rabbitmq`
`ray`
`react`
`reddit`
`redis`
`redpanda`
`redshift`
`rockset`
`rust`
`s3`
`sagemaker`
`salesforce`
`scala`
`scikitlearn`
`scipy`
`scylladb`
`sdf`
`secoda`
`segment`
`sharepoint`
`shell`
`shopify`
`sigma`
`slack`
`sling`
`snowflake`
`snowpark`
`soda`
`spanner`
`spark`
`sql`
`sqlite`
`sqlmesh`
`sqlserver`
`starrocks`
`stepfunction`
`stepfunctions`
`stitch`
`stripe`
`superset`
`tableau`
`teams`
`tecton`
`tensorflow`
`thoughtspot`
`trino`
`twilio`
`twitter`
`typescript`
`vercel`
`wandb`
`x`
`xgboost`
`youtube`
`bronze`
`silver`
`gold`
`dag`
`task`
`table`
`view`
`dataset`
`semanticmodel`
`source`
`seed`
`file`
`dashboard`
`report`
`notebook`
`workbook`
`csv`
`pdf`
`yaml`

Requesting additional icons#

The kinds icon pack is open source and anyone can contribute new icons to the public repo or request a new icon by filing an issue. Custom icons per deployment are not currently supported.