Adding Triggers
Triggers allow you to automate the execution of your Agent's task, without manually telling the Agent to "Go!"
- Any user can create a Trigger for a Public or Chat-only Agent.
- No limit exists on the number of Triggers that can be set up.
- Other users will see both the Trigger setup and the output of that Trigger in the Conversation History.
Available Trigger Types
- Time-Based Triggers — Schedule your Agent to execute on a Daily, Weekly, Monthly, or Interval schedule.
- Incoming Webhook Triggers — Kicks your Agent off when a particular event occurs in an external app.
Configuring Time-Based Triggers
Schedule
- Hourly - The Agent runs its task every X amount of decided upon hours.
- Daily — The Agent runs its task every day at the specified time.
- Weekly — The Agent runs on one or multiple selected days per week at the specified time.
- Monthly (new) — The Agent runs once a month on a date you choose. Pick any date from 1–31 or select Last day of the month.
- Dates 29, 30, and 31 genuinely skip short months — for example, a 31st schedule won't run in February or April.
- When you select one of these dates, an Upcoming Runs list automatically expands so you can see which months would be skipped before you save.
- Interval (new) — The Agent repeats on a fixed cadence anywhere from every 15 minutes up to every 28 days.
- Timezone — The schedule follows the timezone set in your NinjaCat Organization Settings.
Trigger Access Scope
When a Trigger runs, it acts on behalf of a specific identity. You control which identity it uses.
- Your access (default) — The Trigger run sees exactly what you (the creator) can see. Access follows the creator — if your permissions change, the Trigger's access changes with them.
- Organization-wide — The Trigger run sees everything the organization can reach. It belongs to the org rather than an individual, so it keeps running correctly even after someone leaves. Only a full-access user can create or change an org-wide Trigger.
Org-wide Triggers are flagged for review in the Trigger list. You can narrow one to a single person's access or confirm org-wide to clear the flag.
⚠️ Action Required: Review Access on Your Existing TriggersAs part of this update, every existing Trigger has been defaulted to Organization-wide access and flagged "Needs Attention."
- Existing Triggers continue running exactly as configured — nothing is broken.
- Each existing Trigger is now set to Organization-wide access, meaning it can see everything the org can reach.
- You must review each flagged Trigger and either:
- (a) Confirm that Organization-wide access is correct, or
- (b) Change it to Your access (user-scoped).
- Reviewing a Trigger clears its "Needs Attention" flag.
Upcoming Runs Preview
Before you save a Time-Based Trigger, the next 5 scheduled runs are listed, computed in your org's timezone. Use this to confirm the schedule looks right — especially useful for Monthly triggers with dates that may skip certain months.
Status
- On — The Trigger is active and runs according to its schedule.
- Paused - The Trigger is inactive. Both Time-Based and Webhook Triggers can be paused and resumed at any time without losing any of their settings.
Message (Trigger Command)
The message field is what gets sent to the Agent at the scheduled time, instructing it to execute its task based on the original Instructions set up for that Agent.
Example Messages:
- "Do your thing, but only for these two accounts: Athleta and Nike."
- "Do your thing, but only for accounts owned by Julie Brown."
- (Default message if left blank: "Go!" — The Agent will follow its configured prompt as-is.)
Notifications (Optional)
Enable this feature if you'd like an email notification sent to one or more recipients when the Agent finishes its task.
- The recipient list supports comma or semicolon separation (e.g.
[email protected], [email protected]or[email protected]; [email protected]). - Turning notifications off keeps the recipient list intact — re-enabling later doesn't require retyping addresses.
Run History & Statuses
Each Trigger maintains a filterable run record you can review at any time. Each run entry shows:
| Status | Meaning |
|---|---|
| Scheduled | The run is queued and waiting for its trigger time. |
| Running | The Agent is actively executing. |
| Completed | The run finished successfully, with duration and finish time recorded. |
| Failed | The run encountered an error; details are included in the record. |
At a Glance & the Org-Wide Automation List
- Every Trigger shows its last run and next run inline in the Trigger list — no need to open each one to check status.
- The org-wide Automation table gathers all Triggers across the organization in one place. You can filter by:
- Run status (Scheduled, Running, Completed, Failed)
- Triggers that need an access review ("Needs Attention" flag)
Agent Self-Follow-Ups
An Agent can schedule a single follow-up from within a conversation — for example, "check again in half an hour" — and then resume in the same conversation thread when that time arrives.
Limits:
- The follow-up time must be between 1 minute and 30 days from the current run.
- A chain of follow-ups can include up to 5 links.
- An Agent can have up to 10 follow-ups waiting at the same time.
- A follow-up cannot widen its own access — it inherits the access scope of the Trigger or conversation that created it.
Create Scheduled Triggers with Agent Builder
You can describe a schedule in Agent Builder and have it populate the Trigger form automatically.
- Describe what you want in plain language — for example, "run every Monday morning and email me the results" — and Agent Builder will fill in the schedule, message, and notification fields accordingly.
- Nothing saves until you save the agent. You can review and adjust the trigger configuration before committing.
- Agent Builder handles scheduled (Time-Based) Triggers only — it does not create or modify Webhook Triggers, and it leaves any existing webhooks untouched.
Configuring Webhook Triggers
- Name — A name for the Webhook Trigger. This will also appear in the Output listing in the Conversation History.
- URL — The URL to send the webhook to. Copy and save this in a secure location, then add it to the external source that supports outbound webhooks.
- Note: Not all sources/apps support outbound webhooks — confirm this with their documentation and support team before proceeding.
- Instructions — Write instructions the Trigger will use to kick off the Agent. If the prompt is already sufficient, just type "Go!" The body of your webhook POST request will be appended to the end of the prompt.
Viewing Trigger Outputs
Once a Trigger has executed, its Triggered Run will appear in the Agent's Conversation History, including the date and time it ran. The Conversation History can be filtered to show only Triggered Runs, only your Conversations, or All. Clicking into a Triggered Run displays the output in the conversation space for review.
If Notifications have been enabled, an email is sent to the listed recipients when the Agent finishes. That email contains a button that directs the recipient back to the output in the Conversation History.

Updated 21 days ago