Connect your AI assistant
Add the Chytapay MCP server to Claude, Claude Code, Codex, Cursor or VS Code so your assistant answers from this documentation.
Chytapay publishes an MCP server with the public documentation of the Integration API and the widget. Connect it to your AI assistant and ask how to integrate without leaving your editor.
What it is
It is a read-only server. It answers with the pages of this documentation and the rules for agents. It asks for no login, runs no operations and sees nobody's data: it does not create charges, look up merchants or read credentials.
It exposes two tools:
| Tool | What it does |
|---|---|
search_ |
Searches the documentation and returns the relevant passages. |
get_ |
Returns a full page by its id. |
The URL
https://mcp.chytapay.com.ar/mcpIt is an MCP server over HTTP, with no authentication. If your client asks for the transport type, choose HTTP.
Add it to your client
Claude Code
claude mcp add --transport http chytapay https://mcp.chytapay.com.ar/mcpCursor
{
"mcpServers": {
"chytapay": {
"url": "https://mcp.chytapay.com.ar/mcp"
}
}
}VS Code
{
"servers": {
"chytapay": {
"type": "http",
"url": "https://mcp.chytapay.com.ar/mcp"
}
}
}Codex
[mcp_servers.chytapay]
url = "https://mcp.chytapay.com.ar/mcp"claude.ai
https://mcp.chytapay.com.ar/mcp- claude.ai and Claude Desktop (custom connector). Go to Customize > Connectors, click Add and choose Add custom connector. Enter the name
Chytapayand the URL from the claude.ai tab, and under Authentication choose No sign in. To use it in a chat, click +, Connectors and turn it on. Custom connectors are available on all plans, and on Team and Enterprise an organization owner adds them from Organization settings > Connectors. - Claude Code (command). Run the command in your terminal. Add
--scope projectto share it with your team in the repository's.mcp.json. - Codex. Save it in
~/.codex/config.toml(or in a trusted project's.codex/config.toml). It is a remote HTTP server and Codex supports it directly, no bridge needed.codex mcp listshows that it is configured. - Cursor. Save it in
.cursor/mcp.jsonfor one project, or in~/.cursor/mcp.jsonfor all of them. - VS Code. Save it in the project's
.vscode/mcp.json.
If "Add to Cursor" or "Add to VS Code" does not open the editor, copy the MCP URL from the section The URL and add it by hand as explained above.
Check that it works
Ask your assistant something only this documentation answers, for example: "According to the Chytapay docs, what is the daily WhatsApp message cap per integration?". With the server connected, the assistant calls search_chytapay_docs and cites Operations and errors.
In Claude Code, /mcp shows the server as connected.
Warning: The server only knows the public documentation. To operate your integration (create charges, view merchants) you keep using the Integration API with your credentials.
Ready-made prompts
The server also offers four prompts. They are optional: the user triggers them by hand and the assistant does not use them on its own. Each one gives the assistant the path through the documentation and the rules to follow, so it does not improvise. The prompts are written in Spanish.
integrate_widget. Guide to integrate the checkout widget with the published version and examples.first_sandbox_payment. Guide to create a Sandbox charge, simulate its payment from the Portal and receive the webhook.receive_webhooks. Guide to receive webhooks, with a checklist for when an event does not arrive.prepare_production. Guide to review an integration before requesting and using Production credentials.
In Claude Code, type / and look for chytapay: each one is listed as /chytapay:<name> (MCP). You can also type it in full as /mcp__chytapay__<name>, for example /mcp__chytapay__first_sandbox_payment. This assumes you added the server under the name chytapay, as in the command above.