Sign in
  1. Home
  2. Docs
  3. Management API
DOCS

Management API.

Everything the console does is an HTTP call. Create a key, send it as a bearer token, and drive pipelines from CI or your own tools.

Authenticate

In the console, open API keys and create one. The token starts with pgcdc_ and is shown once — copy it then. Send it on every request as Authorization: Bearer pgcdc_….

A key acts with member rights in its workspace: it can manage pipelines and destinations, but not billing or the organisation. Requests are limited to 50 per second per workspace by default; above that you get a 429.

curl -s https://app.waltail.com/api/pipelines \
  -H "Authorization: Bearer pgcdc_…"

Endpoints

Method and pathWhat it does
POST /api/sources/checkRun the six connection checks against a connection string
POST /api/sources/tablesList tables, with whether each can replicate and its replica identity
POST /api/sinks/checkLive-check a destination type and config
POST /api/pipelinesCreate a pipeline: sets up the publication and slot, adds destinations
GET /api/pipelines, GET /api/pipelines/{id}List and read pipelines with per-destination counters
PATCH /api/pipelines/{id}Rename, or set state running / paused
DELETE /api/pipelines/{id}Drop the slot and publication, then delete
GET /api/pipelines/{id}/tables, PUT …/tablesRead or replace the pipeline's table set
POST /api/pipelines/{id}/backfills/estimateRows and bytes a backfill of the given tables would deliver
POST /api/pipelines/{id}/backfills, GET …/backfillsStart a backfill ({"tables":[…],"confirm":true}); list jobs with per-table progress
POST /api/backfills/{id}/cancelStop a running backfill
POST /api/pipelines/{id}/sinksAdd a destination to a running pipeline
PATCH /api/sinks/{id}, DELETE /api/sinks/{id}Change a destination's name, batch size, rate limit or config; remove it
GET /api/pipelines/{id}/messagesMessage explorer: paged, filter by table, state and time
GET /api/pipelines/{id}/activity, …/trafficCaptured count and latest event; delivered and failed per minute, per destination
GET /api/sinks/{id}/undeliveredParked messages with payload, attempts and last error
GET /api/sinks/{id}/messages/{seq}/historyEvery attempt for one message
POST /api/sinks/{id}/replayRequeue undelivered — {"seq":N} for one, empty body for all
POST /api/sinks/{id}/replay-deliveredRe-deliver a {from,to} range inside the replay window
GET /api/usageEgress by hour, day, week, month or month-to-date, per pipeline and destination
GET /api/planThe workspace's plan, retention and replay window
POST /api/keys, GET /api/keys, DELETE /api/keys/{id}Manage API keys

Example: create a pipeline

curl -s https://app.waltail.com/api/pipelines \
  -H "Authorization: Bearer pgcdc_…" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "orders",
    "source_dsn": "postgres://waltail:…@db.example.com:5432/app",
    "tables": ["public.orders", "public.invoices"],
    "sinks": [{
      "name": "orders-webhook",
      "type": "webhook",
      "config": {"url": "https://example.com/hooks/waltail", "secret": "…"},
      "batch_size": 50
    }]
  }'

The destination type values are the kinds listed on each destination page (webhook, kafka, sqs, sns, eventbridge, kinesis, s3, redis, redis_string, pubsub, elasticsearch, opensearch, typesense, meilisearch, nats, rabbitmq, eventhubs, servicebus, cloudflare_queues), and the config keys are the field names in each page's configuration table.

Start streaming in minutes.