Triggers

Start runs without anyone clicking Run, such as from a webhook.
View as Markdown

Most runs start when a person opens a project and clicks Run. But a playbook can also start on its own — for example, the moment an email arrives or another system sends a signal. This page explains the two settings that control how a playbook can be started, and what an automatic start means for you day to day.

These two switches live on the Starting runs card inside the playbook editor’s Settings tab. For the rest of the editor, see Build a playbook.

The two ways a playbook can start

A playbook has two independent on/off switches that decide who — or what — is allowed to start a run.

SettingWhat it allowsDefault
People can start this playbookA person can start a run by hand from the project’s Playbooks page.On
Start from a webhookAnother system can start a run automatically by sending a signal to RakerOne.Off

These are not either/or. You can leave both on so a playbook can be started by hand and automatically. Turning People can start this playbook off hides the manual Run button, which is useful when a playbook should only ever be triggered by another system.

What “Start from a webhook” means

A webhook is a way for an outside system to start a run automatically, without anyone in RakerOne lifting a finger. The setting’s own description says it plainly: it “lets other systems start this playbook — for example when an email arrives.”

In practice, you connect another tool — your email inbox, an intake form, an accounting system — to RakerOne. When the agreed event happens in that tool, it sends a signal to RakerOne, and a new run of this playbook starts. From there the run behaves exactly like one you’d started by hand: tasks light up, people get their assignments, and Florent gets to work where it’s enabled.

1

Turn the switch on

In the playbook editor, open the Settings tab and switch on Start from a webhook on the Starting runs card.

2

Connect the outside system

Connecting the other tool to RakerOne is a technical, one-time setup. The system that calls RakerOne needs an API key to identify itself. Hand this part to whoever sets up integrations — once it’s wired up, you don’t touch it again.

3

Let it run

From then on, each agreed event starts a fresh run on its own. You’ll see new runs appear under Recent work on the project’s Playbooks page, and the people assigned to the first tasks get notified as usual.

Setting up the connection itself — and the credentials it uses — is handled with an API key. See API keys.

What an automatic start means for you

An automatically started run is a normal run. The only difference is that nobody clicked Run — so there are a couple of things worth knowing.

  • You’ll find new runs under Recent work. Because no one launched them by hand, automatic runs show up in the project’s Recent work list rather than being something you kicked off. Check there to see what’s come in.
  • You still do and approve the work. Automatic starts only handle the starting. Everything after that follows the usual rules: people claim and complete their tasks, and a human still approves anything Florent or an automated step drafts. AI drafts, humans approve — an automatic start never changes that.
  • Assignments still apply. Whoever is assigned to the first tasks gets notified the moment the run begins, the same as a hand-started run.

An automatic start does not skip approvals or assignments. It only removes the manual click that begins the run. Every sign-off and hand-off still happens as designed.

Where to go next