Automate itLesson 1 of 5

Everything that starts it: the Triggers tab

Every script and flow has a Triggers tab beside the code. It gathers everything that can start the item without you pressing Run — its API endpoint, its schedules and its event triggers — plus the two limits that protect it. In this lesson you read the tab top to bottom, change the endpoint with Configure, and see both limits do their job.

8 min readBefore this: Publish the tested version, Call it from another systemOpen in Waiso

What you will learn

Read the API card and copy a Sync or Async curl that works as shown
Change method, mode, authentication, parameters and limits with Configure
Add and pause schedules without leaving the editor
Cap calls per minute and runs at the same time

One tab, every way in

Open a script in the editor and pick Triggers on the side rail (in a flow, it is one of the tabs on the right). The tab has four parts, always in this order: API, Schedules, Events and Limits. Each part reads and writes the same records as the project pages — the endpoint is the one on the API page, the schedules are the ones on the Schedules page — so a change made in one place shows in the other at once.

docs-demoCodehello.pySavePublish

hello.py

1def main(name: str = "world"):2    print(f"saying hello to {name}")3    return {"message": f"Hello, {name}!"}

Triggers1

APIOpen in API page

Another system starts this script with an HTTP request — copy the URL or the curl below.

POST/exec/p_58e3011e/hello

ModeSyncAuthBearerLimit60/minStatus

Call itSyncAsync

1curl -X POST "$WAISO_URL/exec/p_58e3011e/hello" \2  -H "Authorization: Bearer $WAISO_TOKEN" \3  -H "Content-Type: application/json" \4  -d '{"name":"world"}'

SchedulesOpen in Schedules

morning_helloWeekdays at 08:00Active

Add schedule

Events

Run this script when something happens elsewhere: an e-mail arrives, a database row changes, an MQTT message comes in.Add event trigger

Limits

Calls per minute60Runs at the same timePlan limit

The API card on top with its URL, the Call it curl, then Schedules, Events and Limits.

Every trigger runs the published version. Until the item is published, the tab says Triggers run the published version. Publish this script to call it by URL or on a schedule. and its add buttons stay disabled with the reason on hover. After a publish, if you keep editing, a line warns You have changes that are not published — triggers still run the published version.

The API card and its curl

Turn the endpoint on
On an item that is Not on the API yet, press Turn on endpoint. The card then shows the method chip, the URL with its path in bold, the columns Mode, Auth, Limit and Status, and Open in API page, which opens the API page focused on this endpoint.
Copy the curl
Under Call it, a Sync | Async switch picks which example is shown, then Copy curl. The switch is display only — it never changes the endpoint, because switching a live endpoint’s mode would break its callers. When the example differs from the endpoint, a line says This endpoint is Synchronous — change it in Configure.
Set the token
The curl reads the token from $WAISO_TOKEN (Set $WAISO_TOKEN to an API token.). Create token opens the API page’s settings with a new token for this endpoint.

Sync waits and answers with the output (best under ~30 s). Async answers at once with a runId; poll it until it finishes. A flow that stops at an Approval step answers 202 with the run while it waits, instead of holding the call open. The full contract — status codes, the poll URL, token errors — is in API endpoints.

Configure: the endpoint’s settings

Configure opens the same dialog as the API page. Its sections are listed on the left; each one is a click away and Save changes writes them together. The dialog’s labels are in English in both languages of the app.

Section	What you choose
Endpoint URL	The path after the project id. Letters, numbers, `-`, `_` and `/` only.
HTTP Method	GET, POST (the default), PUT or DELETE.
Execution Mode	Synchronous (waits and returns the output) or Asynchronous (returns a `runId` to poll).
Authentication	Protected (an API token is required — recommended) or Public (no token).
Parameter Passing	Only when the script has parameters: Request body (JSON) or Query string.
Limits	Calls per minute and Runs at the same time — the same two fields as the tab.
Active Tokens	Only for a protected endpoint: the tokens that can call it.

Schedules and events

Schedules lists every schedule that runs this item, as the Schedules page shows them, with its next run (a paused one shows “—”). Add schedule opens the schedule wizard already pointed at this item; Pause schedule and Resume schedule work on each row; Open in Schedules goes to the full page. Events holds the triggers that start the item when an e-mail arrives, a database row changes or an MQTT message comes in — see Event triggers.

Each schedule row has its own pause switch.

The two limits

Protect it from too many calls or too many runs at once. Both fields are typed as plain whole numbers and saved when you leave the field or press Enter (the help line says Saved for a moment). Both are empty by default — then only your plan’s limit applies.

Limit	Range	Over the limit
Calls per minute (endpoint)	1–100,000	HTTP `429`, a `Retry-After` header and `{"code": "rate_limited", "retryAfter": n}`; no run is created.
Runs at the same time (item)	1–1,000	Extra runs wait in the queue and start when one finishes; they never fail.

Calls are counted per endpoint in each clock minute, whichever token makes them; retryAfter is the seconds left until the next minute (1–60).
The call limit needs the endpoint on (Turn the endpoint on to limit its calls.). The runs limit applies to every run of the item: API calls, schedules, events and runs from the editor.
A refused call is still recorded: it shows in the API page’s usage as rate limited.

ExerciseMake the second run waitOpen in Waiso

See Runs at the same time hold a run in the queue instead of failing it.

Before you start

A project where you can create and publish a script

Steps

1Create a script slow-report with the solution’s code and publish it.
2In Triggers → Limits, type 1 in Runs at the same time and press Enter.
3Turn the endpoint on, open Configure, pick Asynchronous under Execution Mode, press Save changes, then Copy curl.
4Run the curl twice in a row, then open Runs.
5Clear the field when you are done.

Expected result

Two slow-report rows: the first Running, the second Queued; about 20 seconds later the second starts, and both end Success with {"slept": 20}.

Hint

An asynchronous endpoint answers at once with a runId, so the second call starts while the first still runs. The Sync | Async switch beside the curl only changes the example.

Show solution

The limit is on the item, not on the endpoint: the second run waited for a free slot and was never refused. Calls per minute would have refused it with 429 instead.

slow-report.py

import time

def main(seconds: int = 20):
    time.sleep(seconds)
    return {"slept": seconds}

Check yourself · 3 questions

You switch the curl example to Async. Did the endpoint change?

No — the switch only changes the example. The mode changes in Configure.

A caller passes Calls per minute. What does it get, and does a run start?

HTTP 429 with Retry-After; no run starts.

Where else can you edit the schedules shown here?

On the Schedules page — both read the same schedules.

1 / 5