Build AI agents in HighLevel by describing the work you need in plain language. Connect Knowledge Bases, configure CRM actions and triggers, and test agents before publishing. Use Computer for dedicated agent environments, parallel browser tasks, secure website authentication, and smoother Take Control interactions.

TABLE OF CONTENTS

• What Is the Managed Agents Builder?
• Key Benefits of Managed Agents
• Prerequisites and Limitations
• Build Agents Using Natural Language
• Intelligent CRM Agents
• Computer for Managed Agents
• Event-Driven Automation
• How to Set Up Managed Agents
• Frequently Asked Questions
• Related Articles

What Is the Managed Agents Builder?

The Managed Agents builder, formerly called Super Agents, creates AI agents from conversational instructions instead of manually configured flowcharts. Agents use the Agent Studio engine to answer questions, work with Knowledge Bases, and perform CRM actions. Each agent follows a draft, test, and publish lifecycle.

Managed Agents also offers a Unified Workspace that brings agent discovery, testing, conversations, activity, and configuration into one screen. The Unified Workspace is currently available through Labs. Locations without the Labs feature enabled continue using the existing Managed Agents interface.

Key Benefits of Managed Agents

Combine instructions, business knowledge, and automation tools to reduce repetitive work while keeping agent behavior aligned with your goals.

• Natural-language setup: Describe the agent's role and behavior instead of manually wiring workflow nodes.
  
  
• CRM automation: Update records, trigger workflows, and respond to configured events.
  
  
• Knowledge-based responses: Connect approved business information through HighLevel Knowledge Bases.
  
  
• Concurrent browser work: Run browser tasks in parallel using dedicated computer environments and secure website credentials.
  
  
• Testing before deployment: Review responses and actions in the testing panel before connecting the agent to live activity.

• Centralized agent management: Find, test, monitor, configure, and publish Managed Agents from one workspace instead of moving between separate pages.

Prerequisites and Limitations

Confirm access and connected resources before configuring an agent so it can use the information and tools required for its tasks.

Prerequisites

• Your user must have permission to access and manage Managed Agents.
  
  
• Configure and test any Knowledge Bases, workflows, or calendars the agent needs.
  
  
• For Facebook lead form triggers, connect the Facebook Page for lead sync and have a Page administrator accept Meta's Lead Ads Terms of Service.
  
  

Limitations

• Available channels and events depend on the triggers and integrations configured in the account.
  
  
• The Unified Workspace is available through Labs. Locations without that Labs feature enabled continue using the existing Managed Agents interface.

Build Agents Using Natural Language

A clear prompt gives the builder the context it needs to create a useful draft. Describe the agent's role, supported channels, Knowledge Bases, permitted CRM actions, and behavior. You can also start from a prebuilt use case.

Example Prompt: “Create a Social Media Assistant that writes and schedules on-brand posts for Facebook, Instagram, and LinkedIn. Use our Marketing Knowledge Base for product details and follow our brand voice.”

The builder generates a draft with proposed behavior, triggers, and actions. Refine it with follow-up instructions, such as “Make the tone more formal” or “Limit posts to three per week per platform.”

You can also describe Facebook lead form or survey triggers in natural language, then select the available Pages, Lead Forms, or surveys. You do not need to look up their IDs manually.

Intelligent CRM Agents

Managed Agents combine AI reasoning with business information and configured tools to answer questions and complete tasks inside HighLevel.

• Knowledge Base answers: Use documents, rich-text content, tables, CSVs, and crawled web content to ground responses in your business information.
  
  
• CRM actions: Update contact fields and tags, trigger workflows, and create tasks, notes, or appointments according to the agent's instructions and available tools.
  
  
• Media generation: Generate images, audio, and video; edit existing PNG images; and save generated assets to Media Storage. Follow-up requests can refine images within the same conversation.
  
  

Common use cases include social media assistance, Knowledge Base support, community moderation, email campaigns, and routine CRM tasks. Agent Studio also supports external API calls and other available tools.

Computer for Managed Agents

Computer lets Managed Agents interact with websites and perform browser-based tasks. Dedicated environments, parallel execution, secure authentication, and improved browser controls help tasks progress with less waiting.

Dedicated Computer Environments

Each Managed Agent has its own computer environment. Multiple agents can work independently without interfering with each other's environments, improving isolation, performance, and execution reliability.

Parallel Browser Conversations

A single Managed Agent can run 8–10 browser conversations in parallel, with each conversation using its own dedicated browser tab. Improved concurrency handling reduces interference and lets browser automations progress without waiting for another conversation to finish.

Secure Credentials with HighLevel Keychain

HighLevel Keychain securely stores website credentials so agents can authenticate without exposing sensitive login information. This supports browser tasks that require access to authenticated websites.

Faster Interactions and Improved Take Control

The refreshed Computer interface provides clearer visibility into ongoing browser activity and faster, more responsive browser interactions. Take Control provides smoother transitions between agent-controlled and user-controlled browser activity.

Event-Driven Automation

Triggers start an agent when a configured event occurs, reducing the need to begin each task manually. When media generation is enabled, triggered runs can also create media without an active chat session.

• Tag Changed: Run an agent when a contact gains or loses a specified tag.
  
  
• Form Submitted: Run an agent after a specified form submission.
  
  
• Chat Interface: Start agent interactions through supported chat channels.
  
  
• Facebook Lead Form: Run an agent when a contact submits a lead form from a selected Facebook Page. The agent receives the contact and submitted answers.
  
  
• Survey Submitted: Run an agent when any selected survey receives a response. The agent receives the survey name and submitted answers.
  
  

Other supported events can include schedules, appointments, and contact or opportunity events. A form's file-upload field can also pass a submitted PNG image to the agent as part of the trigger context.

Managed Agents Unified Workspace

The Unified Workspace brings agent discovery, conversations, testing, activity, and configuration together so you can manage common tasks without repeatedly leaving the workspace.

• Agent List: Search, filter, sort, and switch agents. Review their status, descriptions, and recent updates.
  
  
• Agent Workspace: Chat with an agent, test behavior, and continue previous conversations.
  
  
• Agent Details: Manage settings, publishing, activity, and available actions from a collapsible panel.
  
  

How to Set Up Managed Agents

Create a draft, configure its tools and triggers, and test its responses before publishing it for live events.

Step 1: Access Managed Agents

Start in the sub-account where the agent will use your business information and automations.

1. Open your HighLevel Sub-Account.
   
   
2. Click AI Agents.
   
   
3. Select Agent Studio.
   
   
4. Open the Managed Agents tab.
   
   

Step 2: Create an Agent

Give the builder a clear goal and boundaries so the generated draft reflects the work you intend to automate.

1. Click + Create Agent.
   
   
2. Describe the agent's goal, channels, Knowledge Bases, and permitted CRM actions, or choose a prebuilt use case.
   
   
3. Submit the description to generate the draft agent.
   
   

Step 3: Review the Agent and Configure Triggers

Check the generated behavior and connected tools before deciding which events should start the agent. Use the chat-based editing sidebar to refine its tone, rules, and actions. You can configure multiple supported triggers for one agent.

Configure a Facebook Lead Form Trigger

Use this trigger to respond to leads from a connected Facebook Page. Select all Lead Forms on the Page or limit the trigger to specific forms.

1. From the Sub-Account view, click AI Agents, select Agent Studio, then select Managed Agents.
   
   
2. Open the agent you want to configure.
   
   
3. Open Triggers and click + Add Trigger.
   
   
4. Under Communications, select Facebook lead form.
   
   
5. Select the Facebook Page. Keep the selection at the Page level for all its Lead Forms, or select specific Lead Forms.
   
   
6. Use search when needed. Lead Form search is limited to the selected Page.
   
   
7. Publish the agent when its configuration and testing are complete.
   
   

Configure a Survey Submitted Trigger

Use this trigger to process responses from one or more selected surveys. A response to any selected survey starts the agent.

1. From the Sub-Account view, click AI Agents, select Agent Studio, then select Managed Agents.
   
   
2. Open the agent you want to configure.
   
   
3. Open Triggers and click + Add Trigger.
   
   
4. Under Forms & Surveys, select Survey submitted.
   
   
5. Select one or more surveys. Use search when needed.
   
   
6. Publish the agent when its configuration and testing are complete.
   
   

When you publish, HighLevel checks that the selected Page, Lead Forms, or surveys still exist and identifies invalid selections. For Facebook lead forms, it also validates that the selected Page and form combination can trigger the agent.

Step 4: Configure Media Generation

Enable the media capabilities required for the agent's tasks, then choose the relevant image or audio settings.

1. From the Sub-Account view, click AI Agents, select Agent Studio, then select Managed Agents.
   
   
2. Open the agent and select Capabilities.
   
   
3. Turn on Image generation, Audio generation, or Video generation as needed.
   
   

Image Generation and Editing

• Select Low, Medium, or High image quality. Medium is the default.
  
  
• Specify the image shape in your prompt. Supported ratios are 1:1, 4:5, 2:3, 3:2, 9:16, 3:4, 4:3, 5:4, 16:9, and 21:9.
  
  
• You can also describe the shape as vertical, widescreen, square, or banner. Without a specified shape, the agent generates a square image.
  
  
• Each new generation request creates a new image unless you explicitly ask to edit an existing image.
  
  
• To edit an existing PNG, describe the change. Continue in the same conversation to refine the most recently edited image without providing it again.
  
  
• Generated and edited images remain available in the conversation context for subsequent requests.
  
  

Example: Generate a 9:16 vertical image of a coffee shop for an Instagram story.

The updated image-generation experience improves speed and the rendering of small text and fine details. Image generation remains consistent across supported aspect ratios. Existing settings and prompts continue to work.

Audio Generation

Select a supported voice and use the Tone field to describe how the audio should sound. Write the source text in the language you want the agent to speak.

Video Generation and Storage

Managed Agents generate eight-second videos in a 16:9 aspect ratio. Audio and video are generated separately. Generated media is saved to Media Storage for reuse.

Step 5: Test the Agent

Test representative requests to check the agent's responses, actions, and configured boundaries before live use.

1. Open the agent's testing panel.
   
   
2. Run requests that exercise its Knowledge Base, CRM actions, and configured instructions.
   
   
3. For web-search responses, review Sources and open citations to check the information.
   
   
4. Test a sample asset when media generation is enabled.
   
   
5. Refine the agent through chat-based editing and repeat testing as needed.
   
   

Step 6: Publish the Agent

Publishing makes the configured agent available for live events. After testing, change its status from Draft to Published.

Step 7: Monitor and Refine

Review agent performance in Agent Studio and related reporting to identify improvements. Refine instructions as needed and keep connected Knowledge Bases current. Use the Knowledge Base Retrieval Tester to check whether important content can be retrieved.

Frequently Asked Questions

Q: How is a Managed Agent different from a Flow Based Agent?

A Managed Agent is created through conversational instructions. A Flow Based Agent uses a manually configured visual canvas. Use Managed Agents for new agents; existing Flow Agents can continue running and remain available for editing and management.

Q: What happens if my Knowledge Base content changes after I publish a Managed Agent?

The agent can use updated information on its next run once that content has been indexed. Validate important changes with the Knowledge Base Retrieval Tester before relying on them in live interactions.

Q: How can I prevent a Managed Agent from making unwanted CRM changes?

Define explicit boundaries in the agent's instructions, such as requiring human approval for suggested changes. Test those instructions and review the agent's actions before publishing.

Q: Where can I review Managed Agent pricing and usage allowances?

See AI Product Pricing for Managed Agent plans, included runs, and pay-per-use billing.

Related Articles

• Parallel Conversations in Managed Agents
• Manage Agency User Roles and Permissions in HighLevel
• Sub-Account User Roles and Permissions