orchestra-skills

Use this skill when the user wants to convert an Airflow task that uses AirbyteTriggerSyncOperator or AirbyteSensor (apache-airflow-providers-airbyte) into an equivalent Orchestra pipeline task targeting Airbyte Cloud. Triggers: any mention of migrating, converting, or rewriting Airflow Airbyte tasks to Orchestra; any Airflow DAG code containing AirbyteTriggerSyncOperator or AirbyteSensor pointed at Airbyte Cloud.

Use this skill when the user wants to convert a Dagster integration that syncs Airbyte Cloud — AirbyteCloudResource, build_airbyte_assets, load_airbyte_cloud_asset_specs, or airbyte_assets — into an equivalent Orchestra pipeline task targeting Airbyte Cloud. Triggers: any mention of migrating or rewriting Dagster Airbyte assets/resources to Orchestra; any Dagster code importing from dagster_airbyte pointed at Airbyte Cloud.

Use this skill when the user wants to convert a Prefect task that syncs Airbyte Cloud — AirbyteConnection block, trigger_sync_run_and_wait_for_completion, or AirbyteSyncResult — into an equivalent Orchestra pipeline task targeting Airbyte Cloud. Triggers: any mention of migrating or rewriting Prefect Airbyte tasks to Orchestra; any Prefect flow code importing from prefect_airbyte pointed at Airbyte Cloud.

Use this skill when the user wants to convert an Airflow task that uses AirbyteTriggerSyncOperator or AirbyteSensor pointed at a self-hosted Airbyte Server instance into an equivalent Orchestra pipeline task. Triggers: any mention of migrating or converting Airflow self-hosted Airbyte tasks to Orchestra; Airflow DAG code with AirbyteTriggerSyncOperator using a host/port pointing to an on-prem or self-managed Airbyte deployment.

Use this skill when the user wants to convert a Dagster integration that syncs a self-hosted Airbyte Server (open-source / OSS) instance — AirbyteResource (with host/port) plus build_airbyte_assets — into an equivalent Orchestra pipeline task. Triggers: any mention of migrating self-hosted Airbyte Dagster assets to Orchestra; Dagster code using dagster_airbyte AirbyteResource configured with a host/port pointing to an on-prem or self-managed Airbyte deployment.

Use this skill when the user wants to convert a Prefect task that syncs a self-hosted Airbyte Server — AirbyteConnection block, trigger_sync_run_and_wait_for_completion, or AirbyteSyncResult pointed at a non-Cloud host — into an equivalent Orchestra pipeline task targeting Airbyte Server. Triggers: any mention of migrating or rewriting Prefect Airbyte tasks to Orchestra; any Prefect flow code importing from prefect_airbyte pointed at a self-hosted Airbyte Server.

Use this skill when an Airflow DAG uses on_failure_callback, on_success_callback, on_retry_callback, sla_miss_callback, EmailOperator, PagerDutyEventsHook, or any notification/alerting pattern. Covers all six Orchestra alert destination types: SLACK, EMAIL, PAGER_DUTY, MICROSOFT_TEAMS, WEBHOOK, and DATADOG. Read this skill alongside slack-airflow-to-orchestra for complete notification coverage.

Use this skill when an Airflow DAG contains BashOperator (non-dbt), SSHOperator (non-dbt), WinRMOperator, or KubernetesPodOperator. Triggers: any task using bash_command= that isn't running dbt, any task using ssh_conn_id= that isn't running dbt, any KubernetesPodOperator or DockerOperator. Note: BashOperator or SSHOperator tasks running dbt CLI commands (dbt run/build/test/seed/snapshot) are handled by dbt-core-airflow-to-orchestra instead, even though they SSH into a remote server.

Use this skill when an Airflow DAG contains conditional execution logic: trigger_rule (all_done, one_failed, none_failed, one_success, none_skipped), BranchPythonOperator, ShortCircuitOperator, or LatestOnlyOperator. Triggers: any Airflow task with a non-default trigger_rule; any DAG using branching or short-circuit patterns; any task that should only run based on upstream status rather than simple success.

Use this skill when converting Airflow code that references credentials via Airflow's connection system: any operator/hook kwarg ending in `_conn_id` (conn_id=, aws_conn_id=, snowflake_conn_id=, ssh_conn_id=, slack_conn_id=, gcp_conn_id=, http_conn_id=), any `BaseHook.get_connection(...)` or `Connection` class usage, any `AIRFLOW_CONN_<NAME>` environment variable, or a Secrets Backend (AWS Secrets Manager / Hashicorp Vault) configured for connections. Also covers distinguishing these from `Variable.get()` / `AIRFLOW_VAR_*`, which are non-secret config, not connections. Must be read before finalising any Orchestra YAML with a connection: field converted from an Airflow source.

Use this skill when an Airflow DAG uses TriggerDagRunOperator, ExternalTaskSensor, or SubDagOperator to create cross-DAG dependencies or trigger downstream pipelines. Triggers: any DAG with external_dag_id=, trigger_dag_id=, SubDagOperator, or Airflow Datasets.

Use this skill when converting any Airflow DAG to Orchestra YAML to handle DAG-level configuration. Triggers: any Airflow DAG definition containing schedule_interval, default_args, max_active_runs, catchup, params, start_date, or DAG-level tags. Must be applied before converting individual tasks — it establishes the pipeline root fields that wrap all task conversions.

Use this skill when an Airflow DAG uses dynamic task mapping — task.expand(param=[...]), task.expand(param=XComArg(...)), task.partial(...).expand(...), or expand_kwargs() (Airflow 2.3+) — to create a variable number of task instances at runtime from a list, dict, or an upstream task's output. Triggers: any DAG using .expand(, .partial(, expand_kwargs(, map_index, or task mapping over a list of files/pages/config entries/API results. Orchestra's equivalent is a matrix task group (TaskGroupModel.matrix), not a plain loop — read this before converting any mapped Airflow task to Orchestra YAML.

Use this skill when an Airflow DAG contains sensor operators: S3KeySensor, FileSensor, SqlSensor, ExternalTaskSensor, TimeSensor, or any *Sensor class. Triggers: any DAG with poke_interval, mode='reschedule', or sensor-based waiting patterns. Sensors in Orchestra are first-class pipeline triggers defined in a sensors: block, not pipeline tasks.

Use this skill when an Airflow DAG contains data quality or testing operators: SqlCheckOperator, SQLThresholdCheckOperator, GreatExpectationsOperator, SodaCheckOperator, DbtTestOperator, or any custom SQL assertion task. Triggers: any task with sql= parameter used for validation, any use of Great Expectations or Soda, any dbt test step.

Use this skill when an Airflow DAG uses xcom_push, xcom_pull, or the TaskFlow API @task return values to pass data between tasks. Triggers: any DAG with ti.xcom_push(), context['ti'].xcom_pull(), @task decorated functions that return values, or BranchPythonOperator decisions driven by upstream XCom values.

Create, validate, and remediate Orchestra pipeline definitions, as YAML or OML. Use when asked to build a new pipeline, add tasks to an existing pipeline, fix pipeline validation errors, or author Orchestra workflow definitions from a description. Trigger on phrases like "create a pipeline", "add a dbt task", "write orchestra yaml", "write orchestra oml", "fix validate errors", or when editing `.yml`, `.yaml` or `.oml` files under orchestra/ or similar pipeline directories.

Use this skill when a Dagster project sends notifications: @run_failure_sensor, @run_status_sensor, make_email_on_run_failure_sensor, make_slack_on_run_failure_sensor, PagerDuty/Teams/Datadog resources used in hooks or sensors, or @failure_hook/@success_hook. Covers all six Orchestra alert destination types: SLACK, EMAIL, PAGER_DUTY, MICROSOFT_TEAMS, WEBHOOK, and DATADOG. Read alongside slack-dagster-to-orchestra for complete notification coverage.

Use this skill when a Dagster project contains data quality / testing logic: @asset_check, asset checks with AssetCheckResult, build_metadata_bounds_checks, dbt tests run via dagster-dbt, dagster-pandera / dagstermill expectations, or ExpectationResult. Triggers: any @asset_check decorated function, any AssetCheckSpec, any SQL/threshold assertion run as a check, any use of ExpectationResult or great_expectations/pandera within an asset.

Use this skill when a Dagster job/graph contains conditional execution logic: ops with multiple Out/branching outputs that are conditionally yielded, @op functions that decide which downstream op runs, DynamicOut fan-out used for branching, or conditional asset materialization (AutomationCondition / skip logic). Triggers: any Dagster op that conditionally yields one of several Outs, any graph that runs different downstream ops based on an upstream result, any asset whose execution is gated on a condition.