> ## Documentation Index
> Fetch the complete documentation index at: https://docs.velafi.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Build with AI

> Connect Claude, ChatGPT, Cursor, VS Code and other AI tools to the VelaFi docs through the VelaFi MCP server, so their answers come from the current documentation.

AI assistants are only as good as the information they have. When you ask a general-purpose assistant about the VelaFi API, it may guess endpoint paths, field names or status codes. The VelaFi MCP server gives your assistant direct, read-only access to these docs, so it can look up the current pages before it answers or writes code.

## What the MCP server does

[Model Context Protocol (MCP)](https://modelcontextprotocol.io) is an open standard that lets AI tools connect to external sources. The VelaFi docs MCP server lets an assistant:

* **Search the docs** for guides, API reference pages, webhooks and error codes.
* **Read full pages**, including request and response fields and examples.

Benefits:

* **Answers from the current docs.** The assistant cites the page it used instead of relying on what it learned during training.
* **Fewer wrong endpoints and fields.** Generated code uses the real paths, parameters and enum values.
* **Works where you code.** Ask questions or generate integration code inside Claude, ChatGPT, Cursor, VS Code or your terminal.
* **Every language on this site.** The assistant can search the English, Chinese and Spanish docs.

> [!NOTE]
>
> The MCP server reads only the public documentation on this site. It cannot access your VelaFi account, call the VelaFi API or move funds, and it needs no API key or login. Never paste your API secret or tokens into an AI chat.

## Server URL

```text theme={null}
https://docs.velafi.com/mcp
```

## Set up your AI tool

> [tab] Claude
>
> In Claude on the web or the Claude desktop app:
>
> 1. Open **Settings → Connectors** ([claude.ai/settings/connectors](https://claude.ai/settings/connectors)).
> 2. Select **Add custom connector**.
> 3. Enter a name, for example `VelaFi Docs`, and the URL `https://docs.velafi.com/mcp`.
> 4. Select **Add**.
> 5. In a chat, open the tools menu (**+**) and turn on the VelaFi Docs connector.
>
> Custom connectors may not be available on every Claude plan, and organization admins can restrict them.

> [tab] Claude Code
>
> Run:
>
> ```bash theme={null}
> claude mcp add --transport http velafi-docs https://docs.velafi.com/mcp
> ```
>
> Run `/mcp` inside Claude Code to check that the server is connected.

> [tab] ChatGPT
>
> ChatGPT supports custom MCP connectors through developer mode, on plans where it is available:
>
> 1. Open **Settings → Apps & Connectors → Advanced settings** and turn on **Developer mode**.
> 2. Create a connector with the name `VelaFi Docs` and the URL `https://docs.velafi.com/mcp`. No authentication is needed.
> 3. In a chat, enable the connector from the tools menu.
>
> Menu names can change as ChatGPT updates its interface. If you can't find these options, check OpenAI's help center for "developer mode" and "MCP connectors".

> [tab] Cursor
>
> Open the command palette, run **Open MCP settings**, then select **Add custom MCP** and add:
>
> ```json ~/.cursor/mcp.json theme={null}
> {
>   "mcpServers": {
>     "velafi-docs": {
>       "url": "https://docs.velafi.com/mcp"
>     }
>   }
> }
> ```

> [tab] VS Code
>
> Add the server to `.vscode/mcp.json` in your project:
>
> ```json .vscode/mcp.json theme={null}
> {
>   "servers": {
>     "velafi-docs": {
>       "type": "http",
>       "url": "https://docs.velafi.com/mcp"
>     }
>   }
> }
> ```
>
> Then use it from GitHub Copilot Chat in agent mode.

> [tab] Windsurf
>
> Add the server to your Windsurf MCP configuration:
>
> ```json ~/.codeium/windsurf/mcp_config.json theme={null}
> {
>   "mcpServers": {
>     "velafi-docs": {
>       "serverUrl": "https://docs.velafi.com/mcp"
>     }
>   }
> }
> ```

> [tab] Codex
>
> Run:
>
> ```bash theme={null}
> codex mcp add velafi-docs --url https://docs.velafi.com/mcp
> ```

Other tools that support remote MCP servers over HTTP can connect with the same URL.

## Try it

Once connected, ask your assistant questions such as:

* "Using the VelaFi docs, how do I generate a business token? Show me a Python example."
* "Which fields do I need to create an off-ramp order to Mexico through SPEI?"
* "What does order status 70 mean, and which webhook tells me about it?"
* "Write a handler that verifies the signature of a VelaFi webhook."

Always review generated code and test it in the [sandbox](https://docs.velafi.com/docs/sandbox.md) before you use it in production.

## Agent skills

To go further than lookups, install the [VelaFi agent skills](https://docs.velafi.com/docs/developer-tools/agent-skills.md). They give your assistant step-by-step instructions, rules and self-check scripts for each core use case.

## Other ways to give the docs to an AI

* **Markdown pages.** Add `.md` to any page URL to get a plain Markdown version, for example [`https://docs.velafi.com/docs/quickstart.md`](https://docs.velafi.com/docs/quickstart.md).
* **Page menu.** Use the menu at the top of any page to copy the page as Markdown, open it in ChatGPT or Claude, copy the MCP server URL, or connect the MCP server to Cursor or VS Code in one click.
* **llms.txt.** [`/llms.txt`](https://docs.velafi.com/llms.txt) lists every page with a short description, and [`/llms-full.txt`](https://docs.velafi.com/llms-full.txt) contains the full text of the docs in one file.


## Related topics

- [Agent Skills](https://docs.velafi.com/docs/developer-tools/agent-skills.md)
- [Generate a Token](https://docs.velafi.com/api-reference/generate-token.md)
- [Convert Local Currency to Stablecoins](https://docs.velafi.com/docs/use-cases/convert-to-stablecoins.md)
- [Pay Out Local Currency Funded by Stablecoins](https://docs.velafi.com/docs/use-cases/payouts-funded-by-stablecoins.md)
- [Sandbox](https://docs.velafi.com/docs/sandbox.md)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.