Support for Dependent/Cascading Dropdown Parameters in Trigger DAG UI

### Description When triggering a DAG via the Airflow web UI, there is no way to define **dependent (cascading) relationships** between `Param` dropdowns — where the selected value in one dropdown dynamically filters the available options in another. For example, if a DAG has a `country` param and a `region` param, selecting **"Kenya"** should only show Kenyan regions (`Nandi County`, `Nyandarua`, `Nakuru`), and selecting **"Ethiopia"** should only show Ethiopian regions (`Jimma`, `Arsi`, etc.). Currently, all regions from all countries are shown in a **flat list** regardless of which country is selected. --- ### Current Workaround The only option today is to flatten all child values into a single `enum` list and validate the parent-child pairing at runtime: ```python from airflow.models.param import Param from airflow.models import Variable COUNTRY_REGIONS = Variable.get("country_region_mapping", deserialize_json=True) # { # "Kenya": {"Nandi County": "21802", "Nyandarua": "26727", "Nakuru": "26199"}, # "Ethiopia": {"Jimma": "123", "Arsi": "456"} # } COUNTRY_LIST = list(COUNTRY_REGIONS.keys()) # Flatten ALL regions into one list — no filtering by country ALL_REGIONS = ["All Regions"] for country in COUNTRY_REGIONS: ALL_REGIONS.extend(list(COUNTRY_REGIONS[country].keys())) dag_params = { "country": Param(default="Kenya", enum=COUNTRY_LIST, type="string"), "region": Param( default="All Regions", enum=ALL_REGIONS, # Shows ALL regions from ALL countries description=" Ensure region matches the selected country", ), } ``` > **Problems:** > - Users can select mismatched pairs (e.g. `Kenya` + `Jimma`) — the DAG fails at runtime instead of preventing the mistake in the UI > - The dropdown is cluttered with irrelevant options from other parent values > - Does not scale as more countries/regions are added > - Requires a dedicated validation task in every DAG that uses hierarchical params --- ### Proposed Solution — `depends_on` Att…