Setting Up Agent Studio Triggers for Real-Time Starts

Agent Studio Triggers power real-time, event-driven automation for your Captivation Hub AI agents. When someone submits a form, gets tagged, or sends a chat message, Captivation Hub records the event and starts your agent in the background — keeping pages and APIs fast while the agent responds instantly. The architecture is built for high volume and safe testing, so you can roll out triggers gradually without disrupting live traffic.

This guide walks through what triggers are, why they matter, and how to configure them quickly with best-practice tips.

What are Agent Studio Triggers?

Agent Studio Triggers are event listeners attached to an agent's Start node that automatically launch the agent when a supported platform event occurs. By reacting to native platform events, your agents can respond the moment something happens — without polling, manual launches, or external automation glue.

Key Benefits

• Real-time response: Agents start immediately when an event fires, so customers don't wait.
• Background processing: Heavy work happens off the main thread, keeping the platform fast.
• High-volume ready: The trigger architecture handles bursts of events without dropping work.
• Safe rollout: Test triggers in staging, then publish — existing API-invoked agents continue to run untouched.
• Foundation for growth: Event-driven design sets the stage for additional event types down the road.

Step 1 — Confirm Prerequisites and Access

Make sure you have access to AI Agents → Agent Studio inside the correct sub-account or location. If you need a refresher on the canvas, testing, and publish lifecycle, see our companion article on using AI Agent Studio in Captivation Hub.

Step 2 — Open Agent Studio

From the left sidebar, click AI Agents, then select the Agent Studio tab at the top. This opens the workspace where you create and manage AI agents for the current location.

Step 3 — Create or Open an Agent

Click Create Agent to build a new one, or pick an existing agent from the list to edit it. The list shows Status, Last Updated, and Created On so you can find the right agent quickly.

Step 4 — Add a Start Trigger to the Canvas

On the canvas, attach a Start trigger to the beginning of your flow. Triggers are configured directly on the Start node so the rest of the agent flow can immediately use the event payload.

Trigger Types

Form Submitted Trigger

Starts the agent when a selected form is submitted. The agent receives the mapped form fields, so the flow can branch, route, or respond based on the submission.

Common uses include lead intake, qualification questionnaires, and follow-up automations driven by specific form responses.

Lead Tag Trigger (Added / Removed)

Starts the agent when a selected tag is added to or removed from a contact. The flow gets the contact context — including existing tags and custom fields — so it can react accordingly.

Common uses include VIP onboarding, re-engagement, and churn-prevention agents.

Chat Message Trigger

Starts the agent the instant a new inbound message arrives through supported chat widgets and providers. This is ideal for instant replies, smart routing, or FAQ deflection.

No additional configuration is required — add the trigger and publish. The agent receives message text, sender details, and conversation context for its flow logic.

Step 5 — Select Forms (Form Submitted Trigger)

For the Form Submitted trigger, use the dropdown to select one or more forms — or click Add new if the form doesn't exist yet. The agent will start whenever any of the selected forms is submitted.

Step 6 — Choose Tag Action (Lead Tag Trigger)

For the Lead Tag trigger, set Tag action to either Tag added or Tag removed. This decides whether the agent starts when a tag is applied to a contact or when it's removed.

Step 7 — Validate and Go Live

Once your trigger is configured, walk through these three steps before publishing:

• Test: Run a safe simulation of the selected trigger to see exactly which variables and payload your agent receives. No live contacts are affected. Use this to iterate on prompts and nodes before publishing.
• Save: Store your current changes as a Staging version so you don't lose work and can come back later. Saving doesn't push anything live. It's a good habit to save after each configuration step (e.g., after choosing forms or tags) so you keep a clean working version.
• Publish: Push the current agent and its Start trigger live for this sub-account so real events immediately start the agent. Publish only after a successful Test, and make sure the Start trigger is wired to the first node. If the agent is mapped to other features (like Ask AI), it typically needs to be in Production to be selectable.

Frequently Asked Questions

Do triggers change how my existing API-invoked agents run?
No — triggers are additive. Your existing API flows keep working. Triggers only fire for agents where you've added a Start trigger.

Where do I turn triggers on?
Inside each agent's canvas in AI Agents → Agent Studio. Add a Start trigger, configure it, Test, then Publish.

Can I test before going live?
Yes. Use the canvas Test option to simulate the event and validate variables and responses before publishing.

Which chat sources are supported?
Triggers work with the chat widgets and providers already connected to your sub-account. No extra configuration is required.

What data is available to the agent when a trigger fires?
Form submissions include the mapped form fields. Lead tag events include contact context. Chat message events include message text and conversation metadata.

What if I don't see my form or tag in the dropdown?
Confirm you're in the correct sub-account, then check that the form or tag exists and is active. Newly created items may take a moment to appear in the picker.