TABLE OF CONTENTS
- What is Invoke Agent Studio Agent?
- Key Benefits of Invoke Agent Studio Agent
- Prerequisites & Permissions
- Workflow Action Settings
- Using Agent Output Downstream
- Best Practices & Example Use Cases
- How To Setup Invoke Agent Studio Agent
- Frequently Asked Questions
What is Invoke Agent Studio Agent?
Key Benefits of Invoke Agent Studio Agent
- AI anywhere: Drop the action before, after, or between any other step for context-aware branching.
- Real-time insights: Summarise conversations, predict next steps, or score leads instantly.
- Multi-channel output: Pipe the agent’s response into Email, SMS, WhatsApp, internal notifications, or custom values.
- Unified data: Update contacts, opportunities, custom objects, or Google Sheets with the agent’s structured JSON output.
- Scalable creativity: Generate copy, images, proposals, or tasks on demand—cut production time and costs.
- No limits on logic: Chain agents, Webhooks, API calls, or If/Else conditions to create end-to-end automations.
Prerequisites & Permissions
- An active AI Employee add-on (Agent Studio access).
- An Agent Studio agent published to Production. Draft or Staging versions will not appear in the dropdown.
- A Chat Trigger enabled on the agent so it can start from a message payload.
- Location permissions to edit workflows and AI agents.
Workflow Action Settings
- Agent: Choose any Production agent in the current sub-account.
- Message (optional): Free-text string that will be sent as the agent’s chat message. Use merge fields for dynamic values.
- Input Variables: Map workflow custom values or static text to the agent’s named input variables. Unmapped required variables will halt the action.
- Store Output As: Pick (or create) a custom value to capture the agent’s full JSON response for downstream steps.
Using Agent Output Downstream
- Update Contact » Set a custom field to the agent’s summary.
- If/Else » Branch when agent.result.score ≥ 80.
- Send Email » Insert into the body.
- Webhook » POST the entire JSON to a third-party service for fulfilment.
Best Practices & Example Use Cases
- Keep prompts short and supply structured context via variables.
- Return machine-readable JSON from the agent for reliable branching.
- Wrap long-running external calls inside the agent (API node) instead of after the action.
- Example 1: “Lead Enrichment” Trigger on Form Submitted → Invoke agent to research company → Update Opportunity & Notify rep.
- Example 2: “Conversation Summary” - After Conversation Closed → Invoke agent to summarise chat → Save note & Send recap email.
How To Setup Invoke Agent Studio Agent
Proper setup begins inside Agent Studio before configuring the Workflow action. The agent must be correctly built and published before it becomes available inside Workflows.
Build the Agent in Agent Studio
Navigate to Agent Studio.

- Create a new agent.
- Select Chat Message as the trigger.

- Configure the agent’s instructions and input variables.
- Click Publish to Production.
Only published agents will appear inside the Workflow action dropdown.Add the Workflow Action
Navigate to Automation → Workflows.
Open or create a workflow.

Click + Add Action.
Under Agent Studio, select Invoke Agent Studio Agent.

Configure the Workflow Action
Once you add the Invoke Agent Studio Agent action to your workflow, the configuration panel determines which agent runs and what data is passed to it. Proper configuration ensures the agent executes successfully and returns meaningful output.
Action Name
The Action Name field allows you to rename this workflow step for clarity. This is especially useful if you plan to use multiple agents within the same workflow.
Example:
Lead Message Analyzer
AI Summary Step
Sales Insight Generator
Renaming this field does not affect how the agent runs — it only helps organize your workflow visually.

Select Agent (Required)
In the Select Agent dropdown:
You will see all agents that:
Are published to Production
Use the Chat Message trigger
Select the specific agent you want to execute in this workflow.

If your agent does not appear:
Confirm it is published to Production
Confirm it uses the Chat Message trigger
Message Field (Optional but Recommended)
The Message field allows you to define what the agent should do when the workflow reaches this step. You can provide instructions and include workflow custom values to pass dynamic data to the agent.
Example: Analyze the contact’s last inbound message and provide a short summary along with recommended next actions for the sales team.
Example with personalization: Analyze the most recent message from {{contact.name}} and provide a short summary along with recommended next actions for the sales team.g
Custom values (such as {{contact.name}}) are resolved before the agent executes, meaning the agent receives fully rendered content specific to that contact.

Important Configuration Notes
Any required input variables defined inside the agent must be mapped correctly.
If required inputs are missing, the agent may fail to execute.
The agent runs immediately when the workflow reaches this step.
The response becomes available for downstream use as {{agent_studio_execution.response}}
Configure Inputs and Save
Map required input variables.
Enter optional instructions in the Message field.
Click Save Action.
Frequently Asked Questions
Related Articles
Was this article helpful?
That’s Great!
Thank you for your feedback
Sorry! We couldn't be helpful
Thank you for your feedback
Feedback sent
We appreciate your effort and will try to fix the article