Skip to content
AI Employee

Overview of Flow Agents in Agent Studio

Complete Guide on how to manage, edit, test, and maintain existing flow-based agents in Agent Studio, and when to use SuperAgents for new agent creation.

5 min read1,087 words6 explained imagesUpdated Wed, 8 Jul at 7:19 AM
View as markdownOpen in ClaudeOpen in ChatGPT
Each image has a one line explanation. Switch to full detail for the step it belongs to, the fields and buttons involved, examples and the whole procedure.
On this page
  1. Key Benefits of Agent Studio
  2. Drag-and-Drop Agent Builder
  3. Tool Nodes
  4. LLM Node
  5. MCP Tool Node
  6. API Tool Node
  7. Knowledge Base Tool Node
  8. Web Search Tool Node
  9. Logic & Flow Control
  10. Version Control
  11. How To Use Agent Studio
  12. Frequently Asked Questions
  13. Related Articles

Agent Studio helps GHL Customer Care users manage, update, and optimize existing flow-based AI agents using a visual builder. While SuperAgents is now the recommended experience for creating all new agents, Agent Studio remains available for editing and maintaining existing flow-based agents. This article explains what Agent Studio is, what it can still do, and how to manage your current agents without disrupting live automations.

Important: Agent Studio’s flow-based builder is being replaced by SuperAgents for new agent creation. Existing flow-based agents will continue to work as expected, and any triggers already connected to those agents will keep running. You can still edit, update, and manage existing flow-based agents from your sub-account. To create new agents, enable SuperAgents from Labs and build them there.

Video 1 Open in a new tab

TABLE OF CONTENTS

What is Agent Studio?

Agent Studio is GHL Customer Care’s flow-based AI agent builder for users who already have existing agents built in the visual canvas experience. It allows you to manage agent logic, update nodes, test behavior, control versions, and maintain connected triggers without rebuilding your automations from scratch. 

While new agents should now be created in SuperAgents, Agent Studio remains an important tool for maintaining and improving your current flow-based agents.

Note: Agent Studio is available to Agency Admins/Users and Sub-account Admins/Users. 

Key Benefits of Agent Studio

  • Visual flow management: Update and maintain existing flow-based agents using a drag-and-drop canvas.

  • Flexible node editing: Adjust LLM, API, MCP, Knowledge Base, Web Search, and other supported nodes inside current agents.

  • Controlled updates: Test changes in staging before promoting them to production.
  • Trigger continuity: Keep existing trigger-based automations running while improving agent logic.
  • Safer optimization: Review and refine decision paths, variables, and outputs without rebuilding the entire agent.
  • Programmatic access: Continue using supported API-based actions for eligible production-ready agents.

Drag-and-Drop Agent Builder

The visual builder makes it easier to understand how an existing flow-based agent works and where updates need to be made. This is especially useful when troubleshooting issues, refining responses, or expanding the logic inside an agent that is already in use.

The Agent Studio canvas allows you to:

  • Open and review an existing agent visually.

  • Move through node connections and decision paths.

  • Update prompt logic and downstream actions.

  • Add or revise supported nodes within an existing agent.

  • Maintain a clearer view of how the full agent operates.

Drag-and-Drop Agent Builder (image 1 of 6) What this shows Open and review an existing agent visually. What this shows Illustrates the "Drag-and-Drop Agent Builder" section of "Overview of Flow Agents in Agent Studio". This screenshot appears in the "Drag-and-Drop Agent Builder" section of "Overview of Flow Agents in Agent Studio". The text alongside this image reads: Open and review an existing agent visually. Move through node connections and decision paths. Update prompt logic and downstream actions. Add or revise supported nodes within an existing agent. Maintain a clearer view of how the full agent operates. Immediately after, the guide continues: Tool nodes extend what an existing agent can do by connecting it to data, logic, and external actions. Updating these nodes helps improve how the agent responds, retrieves information, and completes tasks during live interactions. Image 1 of 6 Where to goOpen and review an existing agent visually. Move through node connections and decision paths. Update prompt logic and downstream actions. Add or revise supported nodes within an existing agent. Maintain a clearer view of how the full agent operates. Next stepTool nodes extend what an existing agent can do by connecting it to data, logic, and external actions. Updating these nodes helps improve how the agent responds, retrieves information, and completes tasks during live interactions.
All 5 steps in this procedure
  1. Open and review an existing agent visually.
  2. Move through node connections and decision paths.
  3. Update prompt logic and downstream actions.
  4. Add or revise supported nodes within an existing agent.
  5. Maintain a clearer view of how the full agent operates.

Tool Nodes

Tool nodes extend what an existing agent can do by connecting it to data, logic, and external actions. Updating these nodes helps improve how the agent responds, retrieves information, and completes tasks during live interactions.

LLM Node

The LLM node leverages advanced large language models to generate intelligent, context-aware responses. You can configure prompts, parameters, and output handling to suit your workflows.

MCP Tool Node

The MCP tool node integrates external tools through the Model Context Protocol, extending your agent’s abilities. This is useful for accessing external data sources or triggering actions.

API Tool Node

The API tool node connects directly to third-party services by sending and receiving data via APIs. This allows your agent to interact with outside systems dynamically.

Knowledge Base Tool Node

Attach a knowledge base to your agent so it can respond based on uploaded documents or stored information. This ensures accurate and consistent answers.

Web Search Tool Node

Allow your agent to retrieve the latest information from the internet. This keeps your agent’s responses fresh, relevant, and accurate.

Logic & Flow Control

Logic and flow control features allow your agent to handle complex scenarios and make decisions based on conditions.

  • Branching via edge conditions: Define different paths depending on variable values or input.

  • Variable management: Create global, runtime, or input variables to ensure context-aware and flexible responses.

Version Control

Agent Studio supports structured lifecycle management with Draft, Staging, and Production modes. This ensures that changes are tested before they go live, preventing disruptions in production.

Agent Studio uses version management to help you test changes safely before they affect live conversations.

Staging:Your work-in-progress version. Saving updates Staging without impacting live traffic.

Production: The live version that runs for real contacts. A Production version is created only when you promote a Staging version.

How To Use Agent Studio

Proper setup ensures you get the most from Agent Studio and its features. Follow these steps to get started:

Note: To get a better understanding of Flow Agents, checkout this detailed guide on How to Setup Flow Agents in Agent Studio

  1. Navigate to AI Agents > Agent Studio from your GHL Customer Care dashboard.

  2. Click on Flow Agents to switch to flow agents list.

  3. Click on the name of an existing flow agent to open it.

    Step 3 of 9: Click on the name of an existing flow agent to open it What this shows Step 3 of 9: Click on the name of an existing flow agent to open it. What this shows Shows what to click for step 3 of 9 in the "How To Use Agent Studio" section of "Overview of Flow Agents in Agent Studio". This animation accompanies step 3 of 9 in the "How To Use Agent Studio" section of "Overview of Flow Agents in Agent Studio". At this point in the walkthrough you click on the name of an existing flow agent to open it. The navigation path used here is AI Agents > Agent Studio. The next step is to add tool nodes (LLM, MCP, API, Knowledge Base, or Web Search) to your workflow. Step 3 of 9Path: AI Agents > Agent StudioImage 2 of 6 What to clickClick on the name of an existing flow agent to open it.
    Buttons and menus referenced ClicknameexistingflowagentAI AgentsAgent StudioFlow AgentsAddtoolnodesConnect
    Next stepAdd tool nodes (LLM, MCP, API, Knowledge Base, or Web Search) to your workflow.
    All 9 steps in this procedure
    1. Navigate to AI Agents > Agent Studio from your GHL Customer Care dashboard.
    2. Click on Flow Agents to switch to flow agents list.
    3. Click on the name of an existing flow agent to open it. this image
    4. Add tool nodes (LLM, MCP, API, Knowledge Base, or Web Search) to your workflow.
    5. Connect nodes with edges to define logic paths.
    6. Configure variables to pass data between nodes.
    7. Use the Test Tab to validate your agent’s flow.
    8. Save and Publish your agent to create/update the Staging version. When you are ready to go live, promote the Staging version to Production.
    9. Deploy Agent.

  4. Add tool nodes (LLM, MCP, API, Knowledge Base, or Web Search) to your workflow.

    Step 4 of 9: Add tool nodes (LLM, MCP, API, Knowledge Base, or Web Search) to your… What this shows Step 4 of 9: Add tool nodes (LLM, MCP, API, Knowledge Base, or Web Search) to your workflow. What this shows Shows what this covers for step 4 of 9 in the "How To Use Agent Studio" section of "Overview of Flow Agents in Agent Studio". This animation accompanies step 4 of 9 in the "How To Use Agent Studio" section of "Overview of Flow Agents in Agent Studio". At this point in the walkthrough you add tool nodes (LLM, MCP, API, Knowledge Base, or Web Search) to your workflow. The navigation path used here is AI Agents > Agent Studio. The next step is to connect nodes with edges to define logic paths. Step 4 of 9Path: AI Agents > Agent StudioImage 3 of 6 What this coversAdd tool nodes (LLM, MCP, API, Knowledge Base, or Web Search) to your workflow.
    Buttons and menus referenced AddtoolnodesAI AgentsAgent StudioFlow AgentsClicknameexistingflowagentConnect
    Next stepConnect nodes with edges to define logic paths.
    All 9 steps in this procedure
    1. Navigate to AI Agents > Agent Studio from your GHL Customer Care dashboard.
    2. Click on Flow Agents to switch to flow agents list.
    3. Click on the name of an existing flow agent to open it.
    4. Add tool nodes (LLM, MCP, API, Knowledge Base, or Web Search) to your workflow. this image
    5. Connect nodes with edges to define logic paths.
    6. Configure variables to pass data between nodes.
    7. Use the Test Tab to validate your agent’s flow.
    8. Save and Publish your agent to create/update the Staging version. When you are ready to go live, promote the Staging version to Production.
    9. Deploy Agent.

  5. Connect nodes with edges to define logic paths.

    Step 5 of 9: Connect nodes with edges to define logic paths What this shows Step 5 of 9: Connect nodes with edges to define logic paths. What this shows Shows what this covers for step 5 of 9 in the "How To Use Agent Studio" section of "Overview of Flow Agents in Agent Studio". This animation accompanies step 5 of 9 in the "How To Use Agent Studio" section of "Overview of Flow Agents in Agent Studio". At this point in the walkthrough you connect nodes with edges to define logic paths. The navigation path used here is AI Agents > Agent Studio. The next step is to configure variables to pass data between nodes. Step 5 of 9Path: AI Agents > Agent StudioImage 4 of 6 What this coversConnect nodes with edges to define logic paths.
    Buttons and menus referenced ConnectnodesedgesAI AgentsAgent StudioFlow AgentsClicknameexistingflowagentAdd
    Next stepConfigure variables to pass data between nodes.
    All 9 steps in this procedure
    1. Navigate to AI Agents > Agent Studio from your GHL Customer Care dashboard.
    2. Click on Flow Agents to switch to flow agents list.
    3. Click on the name of an existing flow agent to open it.
    4. Add tool nodes (LLM, MCP, API, Knowledge Base, or Web Search) to your workflow.
    5. Connect nodes with edges to define logic paths. this image
    6. Configure variables to pass data between nodes.
    7. Use the Test Tab to validate your agent’s flow.
    8. Save and Publish your agent to create/update the Staging version. When you are ready to go live, promote the Staging version to Production.
    9. Deploy Agent.

  6. Configure variables to pass data between nodes.

  7. Use the Test Tab to validate your agent’s flow.

    Step 7 of 9: Use the Test Tab to validate your agent’s flow What this shows Step 7 of 9: Use the Test Tab to validate your agent’s flow. What this shows Shows what this covers for step 7 of 9 in the "How To Use Agent Studio" section of "Overview of Flow Agents in Agent Studio". This screenshot accompanies step 7 of 9 in the "How To Use Agent Studio" section of "Overview of Flow Agents in Agent Studio". At this point in the walkthrough you use the Test Tab to validate your agent’s flow. The navigation path used here is AI Agents > Agent Studio. The next step is to save and Publish your agent to create/update the Staging version. When you are ready to go live, promote the Staging version to Production. Step 7 of 9Path: AI Agents > Agent StudioImage 5 of 6 What this coversUse the Test Tab to validate your agent’s flow.
    Buttons and menus referenced Test TabAI AgentsAgent StudioFlow AgentsClicknameexistingflowagentAddtoolnodes
    Next stepSave and Publish your agent to create/update the Staging version. When you are ready to go live, promote the Staging version to Production.
    All 9 steps in this procedure
    1. Navigate to AI Agents > Agent Studio from your GHL Customer Care dashboard.
    2. Click on Flow Agents to switch to flow agents list.
    3. Click on the name of an existing flow agent to open it.
    4. Add tool nodes (LLM, MCP, API, Knowledge Base, or Web Search) to your workflow.
    5. Connect nodes with edges to define logic paths.
    6. Configure variables to pass data between nodes.
    7. Use the Test Tab to validate your agent’s flow. this image
    8. Save and Publish your agent to create/update the Staging version. When you are ready to go live, promote the Staging version to Production.
    9. Deploy Agent.

  8. Save and Publish your agent to create/update the Staging version. When you are ready to go live, promote the Staging version to Production.

  9. Deploy Agent.

    Step 9 of 9: Deploy Agent What this shows Step 9 of 9: Deploy Agent. What this shows Shows what this covers for step 9 of 9 in the "How To Use Agent Studio" section of "Overview of Flow Agents in Agent Studio". This screenshot accompanies step 9 of 9 in the "How To Use Agent Studio" section of "Overview of Flow Agents in Agent Studio". At this point in the walkthrough you deploy Agent. For context, this section explains: Proper setup ensures you get the most from Agent Studio and its features. Follow these steps to get started. The navigation path used here is AI Agents > Agent Studio. Step 9 of 9Path: AI Agents > Agent StudioImage 6 of 6 What this coversDeploy Agent.
    Buttons and menus referenced DeployAI AgentsAgent StudioFlow AgentsClicknameexistingflowagentAddtoolnodes
    All 9 steps in this procedure
    1. Navigate to AI Agents > Agent Studio from your GHL Customer Care dashboard.
    2. Click on Flow Agents to switch to flow agents list.
    3. Click on the name of an existing flow agent to open it.
    4. Add tool nodes (LLM, MCP, API, Knowledge Base, or Web Search) to your workflow.
    5. Connect nodes with edges to define logic paths.
    6. Configure variables to pass data between nodes.
    7. Use the Test Tab to validate your agent’s flow.
    8. Save and Publish your agent to create/update the Staging version. When you are ready to go live, promote the Staging version to Production.
    9. Deploy Agent. this image

Frequently Asked Questions

Q: Can I still create new agents in Agent Studio’s flow-based builder?
No. New agent creation now happens in SuperAgents.
 

Q: What happens to my existing flow-based agents?
Your existing flow-based agents will continue to work as expected. You can still edit, update, and manage them from your sub-account.

Q: Will my existing Agent Studio triggers keep working?
Yes. Existing triggers connected to your current flow-based agents will continue running as expected.

Q: Can I still edit nodes and logic inside my existing agents?
Yes. You can continue using Agent Studio to edit and manage nodes, logic paths, variables, and related configurations for existing flow-based agents.

Q: Where should I create new agents going forward?
Create all new agents in SuperAgents.

Q: Does Agent Studio still support testing and version management for existing agents?
Yes. Existing flow-based agents can still be tested, updated, and promoted through the available versioning workflow.

Q: Should I move all existing flow-based agents to SuperAgents right away?
Existing agents can continue working as they do today. Use SuperAgents for new agent creation, and manage current flow-based agents in Agent Studio unless your team has a separate migration plan.

Was this guide helpful?

Related guides

Agent Studio - Setting Up Triggers for Flow Agents5 min read · 4 imagesHow to Setup Flow Agents in Agent Studio6 min read · 9 imagesAgent Studio - Router Tool (AI & Conditional Router)8 min read · 14 imagesHow to Create and Manage Folders in Agent Studio4 min read · 8 images