Skip to content

↑↓ navigate ↵ open Ctrl↵ new tab esc close

Shortcuts to integrate faster: the documentation inside your AI assistant and the requests ready to test.

The MCP server lets your AI assistant (Claude, Cursor, VS Code, ChatGPT and others) search and read this documentation while you code. Instead of answering from memory, it looks up the right page and uses the real field names, accepted values and error codes.

https://doc.rapidchargeback.com/mcp/

It only reads the public documentation: it asks for no credentials, does not access your Rapid account and does not call the API. The documentation is available in Portuguese, English and Spanish; ask the assistant to use English (the tools take a language parameter).

Terminal window
claude mcp add --transport http rapid-docs https://doc.rapidchargeback.com/mcp/

In ~/.cursor/mcp.json (or .cursor/mcp.json in the project):

{
"mcpServers": {
"rapid-docs": { "url": "https://doc.rapidchargeback.com/mcp/" }
}
}

In .vscode/mcp.json in the project:

{
"servers": {
"rapid-docs": { "type": "http", "url": "https://doc.rapidchargeback.com/mcp/" }
}
}

Under Settings > Connectors > Add custom connector, enter the name Rapid and the URL above.

Any MCP client with HTTP transport works with the same URL.

ToolWhat it does
search_docsSearches by topic, field name, error code, header or endpoint path. Corrects typos
get_pageReads a whole page in Markdown
list_pagesLists every page, in reading order

Example request: “Using the Rapid documentation, write the alert webhook handler in Node.js, validating the signature.”

Every page has the Copy page button, next to the title. It copies the page as Markdown, the format you can paste into an AI assistant without losing tables or code blocks. The menu next to the button opens the same page directly in ChatGPT or Claude.

Each page is also available as Markdown at its own address, replacing the trailing slash with .md:

https://doc.rapidchargeback.com/en/canais/webhook/payload.md

For tools that read the documentation all at once:

FileContent
/en/llms.txtIndex of every page, with a description and a link to the Markdown
/en/llms-full.txtThe whole documentation in a single file

Every example request in this documentation, organized by the menu sections.

  1. Download the collection.
  2. In Postman, click Import and choose the file.
  3. In the collection variables, fill in client_id and client_secret with the credentials from the dashboard (see Authentication). Basic authentication is already set up in the collection.

The IDs in the examples are fictitious (see Canonical examples): replace them with your account’s before sending. The collection is generated from the examples on the pages, so it follows the documentation.