How to Configure AI Agents in Make.com

This step-by-step tutorial explains how to configure AI Agents in make.com, connect them to an AI provider, and prepare prompts that drive reliable, automated workflows.

By following this guide, you will learn how to create your first agent, define its behavior, and structure conversation templates that can be reused across scenarios.

What You Need Before Setting Up a Make.com AI Agent

Before you start, make sure you have the following prerequisites ready so your AI Agent in make.com can run without errors:

• An active make.com account with access to the AI Agents feature
• Access credentials or an API key for your chosen AI provider, if required
• A clear description of the task you want the AI Agent to perform
• Any sample data or past conversations that can help shape the prompts

Having this information prepared will make it easier to set up your agent and avoid interruptions during configuration.

Create a New AI Agent in Make.com

The first major step is to create a new AI Agent inside your make.com workspace. This is where you define the agent’s basic identity and purpose.

1. Open your workspace and navigate to the AI Agents section.

2. Click the button to create or add a new AI Agent.

3. Enter a clear name for the agent so it is easy to recognize in scenarios.

4. Optionally, add a short description that documents what this agent is meant to do.

After saving these details, you can continue to the configuration area where you define the agent’s behavior and connection settings.

Connect Your AI Provider to Make.com

The AI Agent needs an underlying model to process prompts and generate responses. In this part of the configuration, you connect make.com to your chosen AI provider.

1. In the agent configuration page, locate the section for selecting an AI provider.

2. Choose the provider or model from the available list. This can include general-purpose large language models or task-specific models, depending on what make.com supports.

3. If required, add your API key, token, or connection details. Make sure to store them securely and follow provider guidelines.

4. Save the provider settings to confirm that the AI Agent can access the model.

Once the provider is connected, you can move on to setting up prompts, instructions, and behavior rules that guide how the agent works.

Define the Core Role and Instructions of Your Make.com AI Agent

To get consistent responses, you must define the agent’s role and high-level instructions. This usually appears as a system or base prompt that stays active during the entire conversation.

Write the System Instructions for Make.com

The system instructions tell the AI Agent how to behave. Keep them specific so the model understands the boundaries of its role.

• Describe the agent’s role, such as “support assistant”, “data analyst”, or “content generator”.

• State what the agent can and cannot do, including any compliance or style rules.

• Add examples of answer formats, such as numbered lists, bullet points, or JSON when needed.

• Clarify how the agent should handle missing or ambiguous information.

Save these instructions in the system prompt field so they apply to every interaction that flows through this AI Agent in make.com.

Configure Additional Behavior Rules

Many configurations allow optional behavior rules to refine how the AI Agent responds. Depending on the options available in your version of make.com, you may be able to:

• Limit response length or detail level

• Set a specific tone, such as formal, neutral, or friendly

• Define fallback behavior when the agent cannot answer a request

• Specify which languages the agent should use

These settings help you align the agent’s responses with your workflow, branding, and user expectations.

Build Prompt Templates for Your Make.com AI Agent

Prompt templates allow you to reuse structured instructions and placeholders so that every scenario uses the same reliable pattern. This is a cornerstone of working efficiently with an AI Agent in make.com.

Plan the Prompt Structure

Before you create templates, decide how you want to structure each interaction. A good plan might include:

• A short introduction that reminds the agent of its role

• One or more placeholders for dynamic data from your modules, such as user messages or records

• Formatting instructions so that the agent’s responses are easy to parse

• Optional notes about constraints, like word limits or required sections

Mapping this structure in advance helps you keep prompts clean and consistent across scenarios in make.com.

Create and Save Prompt Templates

After planning, you can create the actual templates:

1. Open the prompt template section for your AI Agent.

2. Create a new template and give it a descriptive name.

3. Write the prompt text, inserting placeholders where scenario data will be merged.

4. Document what each placeholder expects, such as user input, ticket data, or CRM fields.

5. Save the template so it can be referenced by modules and scenarios.

Well-documented templates reduce confusion later and enable other team members to reuse the same agent configuration in multiple automation flows.

Test and Refine Your Make.com AI Agent

Testing is essential to ensure your AI Agent behaves as expected when it receives real data from your scenarios in make.com.

Run Sample Conversations

Use test data or controlled examples to run conversations against the configured AI Agent.

• Send several example prompts that mirror real user questions or tasks.

• Check whether the responses follow the system instructions and template structure.

• Look for gaps, such as missing details, incorrect tone, or formatting issues.

If the results are inconsistent, adjust the system instructions or improve the prompt templates, then repeat the tests until the behavior is stable.

Optimize for Reliability and Clarity

As you refine the configuration, aim for predictable and clear outputs. Consider the following adjustments:

• Simplify complex instructions into shorter, direct sentences

• Add explicit examples of desired answers inside the prompts

• Clarify any constraints, such as “answer only with yes or no plus one short explanation”

• Document all changes so anyone configuring make.com scenarios understands how the agent was tuned

Consistent documentation saves time when you extend your automation or share the setup with colleagues.

Use the Make.com AI Agent Inside Scenarios

After configuration and testing, you can connect the AI Agent to modules inside your scenarios.

1. Open a scenario where you want to use the AI Agent.

2. Add the appropriate module that calls the agent.

3. Select the agent you configured and choose which prompt template to use.

4. Map scenario data to the template placeholders, such as incoming messages or record fields.

5. Run the scenario in test mode and review the execution results.

Iterate on the mappings and templates until the scenario produces the desired output reliably.

Further Resources for Make.com AI Configuration

To deepen your understanding of AI Agent configuration and related automation patterns, consult these resources:

• Official how-to instructions on the Make help center, which provide detailed screens and updated options.

• Professional automation and integration guidance from Consultevo, which can help you design scalable workflows using AI and integrations.

By following these steps and refining your prompts over time, you will be able to configure powerful AI Agents in make.com that automate complex tasks, support your users, and streamline data processing across your tools.

Need Help With Make.com?

If you want expert help building, automating, or scaling your Make scenarios, work with ConsultEvo — certified workflow and automation specialists.

Get Help