orchestra-skills

Use this skill when converting Prefect code that references blocks holding credentials (SnowflakeCredentials, AwsCredentials, SlackWebhook, or any other loaded via Block.load()), Secret blocks, or environment-variable-based credentials to Orchestra. Triggers: any Prefect block that stores host/login/token/key, any Secret.load()/SecretStr used for credentials, or environment-specific block configuration. Must be read before finalising any Orchestra YAML that contains a connection: field.

Use this skill when a Prefect flow triggers another deployment using run_deployment(), or when a flow is called as a subflow from a parent @flow. Triggers: any Prefect code with run_deployment(), await run_deployment(), or a @flow function called inside another @flow (subflow pattern).

Use this skill when a Prefect flow passes data between tasks via return values (.result()), task.submit() futures, Prefect Artifacts, or .map() for fan-out. Triggers: any Prefect flow where a downstream task consumes the return value of an upstream task, any .map() usage, any Prefect Artifact that is read by a downstream step, or any conditional branch driven by a task's return value.

Use this skill when converting any Prefect flow to Orchestra YAML to handle flow-level configuration. Triggers: any Prefect flow definition containing @flow, schedule= (CronSchedule, IntervalSchedule, RRuleSchedule), retries=, timeout_seconds=, flow.deploy() or flow.serve() deployment config, flow parameters (typed function args), or flow-level tags. Must be applied before converting individual tasks — it establishes the pipeline root fields that wrap all task conversions.

Use this skill when the user has Prefect Secret blocks, SecretStr fields, SecretDict, or prefect.blocks.system.Secret usage to move into Orchestra. Triggers: any Prefect code using Secret.load(), .get_secret_value(), SecretStr type annotations, or external secret-store blocks (AWS Secrets Manager, Azure Key Vault) storing credentials. Never asks for real secret values — produces a checklist of what to re-enter directly in Orchestra.

Use this skill when a Prefect flow contains data quality or testing logic: @task functions that run SQL assertions, Great Expectations checkpoints called from a @task, Soda check tasks, or threshold checks that raise exceptions on failure. Triggers: any Prefect @task that queries a database to validate data, any use of great-expectations or soda-core inside a Prefect flow, any task that checks row counts, null counts, or data bounds.

Use this skill when the user wants to convert an Airflow task that uses PythonOperator or PythonVirtualenvOperator into an equivalent Orchestra Python pipeline task. Triggers: any mention of migrating or rewriting PythonOperator tasks to Orchestra; Airflow DAG code using PythonOperator, PythonVirtualenvOperator, ExternalPythonOperator, or @task decorator tasks.

Use this skill when the user wants to convert a Dagster @op or @asset that runs arbitrary Python (computation, API calls, pandas, boto3) into an equivalent Orchestra Python pipeline task. Triggers: any mention of migrating or rewriting Dagster ops/assets to Orchestra; plain @op / @asset / @multi_asset functions containing Python logic (not a dedicated integration like dbt/Airbyte/Fivetran).

Use this skill when the user wants to convert a Prefect @task function containing arbitrary Python logic (computation, API calls, pandas, boto3, etc.) into an equivalent Orchestra Python pipeline task. Triggers: any mention of migrating or rewriting Prefect @task functions to Orchestra; Prefect flow code with @task decorators that are not a dedicated integration (not dbt/Airbyte/Fivetran); ShellOperation tasks running Python scripts.

Use this skill when the user wants to convert an Airflow task that uses SlackAPIOperator, SlackWebhookOperator, or SlackAPIPostMessageOperator — or Airflow on_failure_callback / on_success_callback functions that send Slack messages — into an equivalent Orchestra pipeline alert or task. Also covers a plain PythonOperator (or @task) whose function body calls the Slack API directly via slack_sdk/WebClient/chat_postMessage rather than a dedicated Slack operator. Triggers: any mention of migrating or rewriting Slack notification Airflow tasks to Orchestra; Airflow DAG code using SlackAPIOperator, SlackWebhookOperator, SlackAPIPostMessageOperator, callback functions that fire SlackWebhookOperator, or a PythonOperator whose entire job is posting a Slack message via slack_sdk/WebClient.

Use this skill when the user wants to convert Dagster Slack notifications — SlackResource, make_slack_on_run_failure_sensor, make_slack_on_freshness_policy_status_change_sensor, or op success/failure hooks that post to Slack — into an equivalent Orchestra pipeline alert or task. Triggers: any mention of migrating or rewriting Dagster Slack notifications to Orchestra; Dagster code importing from dagster_slack, or @success_hook/@failure_hook functions that send Slack messages.

Use this skill when the user wants to convert Prefect Slack notifications into an equivalent Orchestra pipeline alert or task. This covers both hook-based notifications — SlackWebhook block, @flow on_failure/on_completion hooks, Prefect Automation Slack notifications — AND explicit mid-flow Slack messages, including a plain @task that calls the Slack Web API directly (slack_sdk, WebClient, chat_postMessage) rather than going through prefect_slack. Triggers: any mention of migrating or rewriting Prefect Slack notifications to Orchestra; any Prefect flow code importing from prefect_slack; on_failure= or on_completion= flow arguments posting to Slack; send_incoming_webhook_message(); or any @task whose body's entire purpose is posting a message to Slack (via slack_sdk.WebClient(...).chat_postMessage(...) or similar) and is called explicitly as a step in the @flow body, not just from a callback.

Use this skill when the user wants to convert an Airflow task that uses TableauOperator or TableauRefreshWorkbookOperator (apache-airflow-providers-tableau) into an equivalent Orchestra pipeline task targeting Tableau Cloud. Triggers: any mention of migrating or rewriting Tableau Airflow tasks to Orchestra; Airflow DAG code containing TableauOperator, TableauRefreshWorkbookOperator, or TableauJobStatusSensor.

Diagnose a failed Orchestra pipeline, open a fix PR, validate it on a branch run, then present a human-readable triage summary and STOP for user approval before merging. Use this when the user wants to review the fix before it goes to main — not for fully automated fixes. Trigger on phrases like "triage my pipeline", "show me what's broken", "investigate but don't fix yet", "prepare a fix for review", or when the user explicitly wants a review gate before applying changes. Also trigger when the user describes a symptom in a downstream system ("dashboard looks wrong", "chart is stale", "dbt model has bad data") even if no pipeline error exists — the skill will trace the symptom upstream.