No separate CLI or scripting tool to learn. Every operation is a function you call from SQL.
agol_fdw exposes its operations as plpython3u functions inside the agol_fdw schema. A DBA can run the whole lifecycle from a SQL client or a scheduled job:
introspect_layer() — read an AGOL layer's definition into the catalog.generate_layer_ddl() — emit the DDL for a catalogged layer.materialize_layer() — introspect → DDL → apply → backfill.sync_schema() — detect and apply schema drift.push_outbox() / apply_inbound() — drive sync on demand.sync_now() — push, apply, and reconcile in one call.reconcile_audit() — periodic catch-up sweep.register_webhook() — register an inbound webhook with AGOL.presigned_url() — time-limited access to an attachment in S3.Everything is scriptable, schedulable with pg_cron or an external scheduler, and observable through ordinary SQL queries against the catalog, queues, and logs.
-- full lifecycle from a SQL client:
SELECT agol_fdw.materialize_layer(
layer_url := 'https://services.arcgis.com/.../FeatureServer/0',
layer_name := 'poles',
target_schema := 'data'
);
-- inspect the catalog, queues, and logs — all ordinary tables:
SELECT * FROM agol_fdw.layers;
SELECT * FROM agol_fdw.sync_run_log ORDER BY started_at DESC LIMIT 10;