MCP Server (AI)

Use the Asaas documentation with AI assistants

Asaas provides two resources to make it easier for artificial intelligence tools to use the documentation:

MCP Server: connects compatible assistants to the documentation and the API reference;
LLMs.txt: provides a structured index of the portal's public content.

Connect your assistant to the MCP Server

The Model Context Protocol, or MCP, lets AI tools query external resources in a structured way.

The Asaas MCP Server uses the documentation and the OpenAPI specification to provide information about the available endpoints.

📘
Once this setup is complete, your assistant will be able to query the Asaas documentation and API reference.

What the assistant can look up

Depending on the features available in the tool you use, the assistant may be able to:

search documentation content;
list API endpoints;
look up parameters and schemas;
view request and response structures;
generate integration examples;
help analyze errors;
run requests when the tool is authenticated and allows this type of action.

Configure the MCP Server

Use the following address:

https://docs.asaas.com/mcp

Add this URL to your tool's MCP configuration.

Add to the ~/.cursor/mcp.json file:

{
  "mcpServers": {
    "asaas": {
      "url": "https://docs.asaas.com/mcp"
    }
  }
}

Also see the MCP Server configuration examples.

Protect your API key

The MCP Server is public for querying the documentation. However, authenticated API operations require a valid Asaas key.

⚠️
Never enter your API key directly in the assistant's prompt.
Configure the credential using the secure mechanisms provided by the tool, such as environment variables or secrets storage.

Before allowing requests to be run:

use a Sandbox key during testing;
limit the key's permissions;
avoid providing Production credentials;
review each operation before it runs;
do not allow the tool to expose the key in logs or responses.

See API Keys and Permission management.

Test the configuration

After connecting the server, ask questions such as:

How do I create a customer in Asaas?
Which endpoint creates a subscription?
Which fields are required to tokenize a card?
Which Webhook events are related to charges?

Check that the answers present endpoints, fields, and links consistent with the current documentation.

📘
Review the generated content before using it in your application.
AI assistants may misinterpret information or generate incomplete examples.

Query another branch of the documentation

By default, the MCP Server uses the main published version.

To query another available branch, add the branch parameter:

https://docs.asaas.com/mcp?branch=branch_name

Check the LLMs.txt file

The llms.txt file presents documentation links and descriptions in a structured format for compatible tools.

Access:

https://docs.asaas.com/llms.txt

This file can help AI tools:

find relevant pages;
identify up-to-date content;
use the documentation's terminology;
find related guides and endpoints.

📘
Use of llms.txt depends on each tool's support and configuration.
The file makes the content easier to access, but it does not guarantee that every AI model will consult it automatically.

Report incorrect information

If you find an answer that is inconsistent with the documentation, send the details to:

📧
[email protected]

Whenever possible, provide:

the tool used;
the question sent;
the answer received;
the related page or endpoint;
the information that needs to be corrected.

Next steps

Updated 2 days ago

Did this page help you?