Agent Studio is Captivation Hub's visual environment for building intelligent AI agents. It gives you a flexible, node-based canvas where you can decide how an agent gets triggered, how it processes input, what decisions it makes, and what actions it takes — all in one place.
This guide is meant for users opening Agent Studio for the first time. We'll walk through the structure, the core components, and how everything fits together before you build anything advanced.
A quick note on naming: AI Studio and Agent Studio are two separate AI tools inside Captivation Hub, built for different purposes. AI Studio is a conversational, prompt-driven builder for creating websites and pages quickly. Agent Studio, which is what we're covering here, is a visual builder for AI agents that respond to events, talk with users, and run automations.
What is Agent Studio?
Agent Studio is a visual builder for creating AI-powered agents that can respond to events, interact with users, retrieve information, generate content, and connect to outside systems.
Unlike basic rule-based automations, agents built here can interpret what's happening, make context-aware decisions, route conversations dynamically, and use tools like knowledge search or external APIs.
Each agent is built by connecting nodes on a canvas. Together, those nodes define how the agent starts, how it thinks, and how it finishes its task.
Key Benefits of Agent Studio
- Visual AI builder: Design agent behavior with a drag-and-drop canvas.
- Event-based execution: Trigger agents from CRM events like chat messages, form submissions, or new lead tags.
- Flexible decision logic: Route conversations using AI intent detection or rule-based conditions.
- Built-in knowledge and tools: Connect your Knowledge Base, search the web, or fire off API calls.
- Multi-modal output: Generate text, images, audio, or video — all from a single flow.
- Structured data collection: Capture emails, phone numbers, and user selections during the interaction.
Understanding the Agent Studio Interface
Step 1 — The Agent Studio Dashboard
When you go to AI Agents → Agent Studio, you'll land on a list of every agent in your account.
From the dashboard, you can:
- View existing agents
- Check status (Draft or Published)
- See creation and last-updated dates
- Organize agents into folders
- Create a new agent
- Browse available templates
This is your central management area for AI agents.
Step 2 — Creating a New Agent
From the dashboard, click Create Agent. You'll have two options:
- Start from scratch
- Use a template from the Template Library
Starting from scratch gives you a blank canvas. Templates give you a pre-built structure that you can adjust to fit your use case.
Step 3 — The Agent Template Library
The Template Library is a collection of ready-made agent blueprints designed for common scenarios.
Instead of building from scratch, you can:
- Browse categorized templates
- Preview how each agent is structured
- Install a template into your account
- Modify any part of the flow afterward
Templates usually include predefined triggers, AI nodes, routing logic, and actions. For first-time users, they're often the fastest way to learn how agents are wired together.
Step 4 — The Builder Canvas
When you open or create an agent, you land on the visual builder canvas. This is where you:
- Drag and connect nodes
- Define execution logic
- Configure each step in the agent
Agents are built by connecting nodes left to right, starting with a trigger and ending at an action or completion point. The connections between nodes are what define the agent's real-world behavior.
Step 5 — The Start Trigger
Every agent begins with a Start Trigger. The trigger tells the agent when to run. Common triggers include:
- Chat Message
- Form Submission
- Lead Tag Added
Without a trigger, the agent won't execute. The trigger ensures the agent only fires when its conditions are actually met.
Node Categories
In Agent Studio, everything on the canvas is built using nodes. A node is a single building block that performs one specific function — starting the agent, generating a response, collecting data, or connecting to another system. By connecting nodes together, you define how the agent behaves end-to-end.
AI Agent: The "brain" of the flow. It reads incoming input, understands the context, and generates intelligent responses based on your instructions. For example, if a user asks about pricing, this node interprets the question and replies appropriately.
Sequential: Groups multiple steps and runs them in a fixed order. Useful when actions must happen one after another — for example, collecting an email first, then sending a confirmation.
End Node: Marks where the agent stops. Once the flow reaches this node, the interaction is complete.
Router: A decision point. The agent can choose between paths based on conditions or AI-detected intent — for example, support questions go one way, sales questions go another.
Search Knowledge Base: Pulls answers from your connected Knowledge Base, so the agent can respond with information from your own documentation.
Search Web: Looks up information online when the answer isn't internal — useful for current data or public information.
MCP Server: Connects the agent to supported external tool servers for advanced integrations and specialized actions.
API Call: Sends or receives information from another application — for example, pushing lead data out or pulling customer details in.
Audio Generation: Generates spoken audio responses based on instructions in the flow.
Image Generation: Creates images from text prompts, useful for dynamic visual content.
Text Generation: Produces written content like messages, summaries, or structured output.
Email Address: Collects an email address from the user in a structured format.
Phone Number: Collects a phone number in a standardized format for clean storage.
Single Choice: Presents predefined options for the user to pick from, helping you guide structured decisions.
Text Input: Collects open-ended responses, letting the agent gather longer or unstructured details.
Variables and Global Prompt
At the top of the builder you'll find configuration tools that control how the agent behaves across the entire flow.
Variables
Variables are reusable placeholders that store information. Define a value once — like a business name — and reference it everywhere instead of retyping it.
Global Prompt
The Global Prompt sets overarching instructions that guide the agent's behavior. This is where you define tone, personality, and the rules the agent should follow on every interaction.
Testing and Publishing
Before going live, you can:
- Use the Test feature to simulate execution
- Save changes as a draft
- Publish the agent when you're ready
An agent must be published in order to respond to real-world triggers.
Getting Started with Agent Studio
To create your first agent:
- Navigate to AI Agents → Agent Studio
- Click Create Agent
- Choose to start from scratch or install a template
- Add a Start Trigger
- Drag an AI Agent node onto the canvas and connect it
- Use Test to simulate behavior
- Save and Publish the agent
Frequently Asked Questions
Do agents run automatically?
Only when their configured trigger conditions are met.
How is Agent Studio different from regular workflows?
Workflows follow predefined automation rules. Agents can interpret input and make dynamic decisions using AI.
Do I need to publish an agent for it to work?
Yes. An agent must be published before it can respond to real events.
Can agents integrate with external systems?
Yes — the API Call and MCP Server nodes both support external integrations.
Can agents generate content beyond text?
Yes. Agent Studio supports image, audio, and video generation.
Comments
0 comments
Please sign in to leave a comment.