Skip to main content
Triggers let Twin agents run without manual input — either when something happens in another app or on a defined schedule.

​
Overview

Triggers are how agents move from “on-demand” to fully autonomous. With triggers, an agent can:
  • Start immediately when an external event occurs
  • Run automatically at fixed intervals
  • Be triggered programmatically from your own backend

Event-based triggers

Agents react to incoming events from other apps

Time-based triggers

Agents run on a defined schedule

REST API

Trigger agents from your own backend via API
Both allow Twin to operate continuously in the background.

​
Event-Based Triggers

Event-based triggers launch an agent as soon as something happens in another system — for example, when a message arrives or new data is created.

​
Built-in Event Sources

Twin supports direct connections to common communication and productivity tools. Once an account is connected, events from that tool can be used to start an agent automatically.

Examples

  • New messages or activity in Slack
  • Updates in GDrive or Attio
  • Changes in Intercom or internal tools

​
External App Events

Twin can also react to events from a wide range of third-party services via event platforms and connectors. This makes it possible to trigger agents from thousands of applications, including:

Gmail

New emails, label changes, or draft creation

Google Calendar

Event creation, updates, or deletions

GitHub

Code pushes, new issues, pull requests

Notion

Page or database updates
Also supports HubSpot, Stripe, Airtable, and many more. If your app supports event-based actions, Twin can likely trigger from it.

​
Custom Webhooks

For systems that aren’t supported out of the box, Twin can expose custom webhook endpoints. These endpoints accept structured HTTP requests and can be called from:
  • Internal services
  • Custom applications
  • Scripts
  • Any platform capable of sending JSON payloads
You simply specify what system should call the webhook and what data it will send — Twin handles the rest.

​
Setting Up Event Triggers

Connect the source system

Link the tool or service that will emit the event.

Define the triggering condition

Specify what should start the agent (for example, “a new message is received” or “a record is created”).

Review and activate

Once configured, the trigger becomes active and will start the agent automatically when the condition is met.
You can see and manage active triggers from your agent settings.

​
Scheduled Triggers

Scheduled triggers run agents at predefined times, regardless of external events. This is useful for:

Data Collection

Recurring data gathering tasks

Reports

Daily or weekly automated reports

Monitoring

Regular monitoring tasks

Maintenance

Routine maintenance workflows

​
Creating a Schedule

Verify the agent

Make sure the agent runs correctly when executed manually.

Choose a schedule

Select how often it should run — hourly, daily, weekly, or a custom interval.

Activate the schedule

Once saved, the agent will run automatically based on the defined timing.

Manage schedules

All active schedules can be viewed and adjusted from the scheduling dashboard.

​
REST API Triggers

If you’re a developer, you can trigger Twin agents directly from your own backend using the REST API. This lets you integrate Twin into any existing application or workflow.
Use API triggers when you want to:
  • Run agents in response to events in your own system
  • Integrate Twin into a custom pipeline or CI/CD flow
  • Build a product where Twin agents power backend logic

REST API Reference

See the full API documentation for triggering runs, managing agents, and more

​
Choosing the Right Trigger

Event-based triggers are ideal when speed matters. Scheduled triggers are best for predictable, recurring work. REST API is best for developers integrating Twin into existing systems.

Start Building

Set up your first trigger in the Twin App