Overview
Triggers determine when your agents and workflows start running. They connect agents to external services or time-based schedules, responding to messages, emails, database changes, or recurring events.
How Triggers Work
You create a trigger using three elements:
- Trigger - The external service (Slack, Gmail, Airtable, GitHub) or cron schedule you connect to
- Binding - Configuration that defines what to monitor (which channel, label, or schedule)
- Target - The agent or workflow to execute when triggered
When an external event occurs (like a new Slack message or Airtable record update), the trigger detects it. The binding checks if the event matches your configuration. If it matches, the target executes.
Where You Can Use Triggers
Triggers eliminate manual agent execution by responding to real-time events or running on schedules. Add automation capabilities to your agents respond instantly to customer actions, process data as it arrives, or run tasks on a schedule.
Here are some examples of what you can build:
Customer Support Automation - Run sentiment analysis agents when messages arrive in your Slack channel.
Email Processing - Automatically categorize and route emails with AI-powered workflows.
Database Sync - Process new or updated records from Airtable with data processing agents.
CI/CD Integration - Run code review agents when pull requests are opened on GitHub.
Scheduled Reports - Generate and send daily analytics reports with AI agents at specific times.
Supported Trigger Providers
| Type | Description | Delivery Method |
|---|---|---|
| Slack | Monitor workspace messages and channels | Webhook |
| Gmail | Watch for emails with specific labels | Polling |
| Airtable | Detect record changes in bases and tables | Polling |
| GitHub | Respond to repository events (PRs, issues, commits) | Webhook |
| Schedule | Execute on cron expressions | Time-based |
Next Steps
See usage documentation for step-by-step instructions on creating and managing triggers.