
Using the Trigger Node
The Trigger Node starts a workflow without manual chat input. It listens for time-based schedules, external integration events, or incoming webhooks and then injects mapped data into downstream nodes.

Configuration Panel
The Trigger Node configuration is organized into sections:- Integration
- Schedule
- Webhook
Runs when a connected tool emits an event.
- Source Integration: e.g. Gmail, Calendar, GitHub, Jira.
- Event Type: email_received, new_issue, meeting_created, workflow_failed, etc.

Examples in Practice
Daily KPI Digest
- Type: Schedule (08:00 daily)
- Maps: kpi_revenue, kpi_churn from external analytics endpoint (via later Tool Node)
- Flow: Trigger → Tool Node (fetch KPIs) → LLM Node (summarize) → Output
Meeting Follow-Up
- Type: Integration Event (Calendar: meeting_created)
- Filter: duration >= 30m AND attendee_count > 2
- Maps: title, start_time, attendees
- Flow: Trigger → Delay (1h) → LLM summary → Send recap
Critical Issue Escalation
- Type: Integration Event (Jira: issue_created)
- Filter: priority=High OR labels contains “security”
- Maps: issue_key, summary, priority
- Flow: Trigger → Condition (priority) → LLM (draft Slack alert) → Tool (post Slack)
Webhook: Deployment Report
- Type: Webhook (POST from CI)
- Payload: serviceName, version, status, metrics
- Flow: Trigger → LLM (generate release note) → Store in Docs
GitHub PR Watcher
- Type: Integration Event (GitHub: pull_request_opened)
- Filter: base_branch=main AND lines_changed > 500
- Flow: Trigger → LLM (summarize changes) → Tool (request reviewers)
Best Practices
| Goal | Tip |
|---|---|
| Noise reduction | Use filters and dedup window |
| Reliability | Set retry with exponential backoff |
| Clarity | Name mapped variables consistently |
| Safety | Start read-only; add write nodes later |
| Scaling | Monitor run volume; adjust caps |
Troubleshooting
| Issue | Fix |
|---|---|
| Trigger not firing | Confirm enabled, time zone, cron syntax |
| Duplicate runs | Adjust dedup window or filters |
| Missing payload field | Verify mapping and incoming field name |
| Webhook rejected | Check secret, content-type, schema |
| Excess volume | Lower max runs per interval or refine filters |
Sample Flow
Trigger Node configured. Workflow will run automatically when conditions are met.


