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

# Building with AI

Integrate Sezzle without leaving your editor. Our MCP server lets an AI agent search these docs, and our skills walk it through building and testing an integration. Every page is also available as markdown for any large language model (LLM).

## MCP Server

Sezzle has an <Tooltip tip="An MCP server implements the Model Context Protocol, a standard for connecting AI models to external tools and data.">MCP server</Tooltip> that lets you search this documentation in natural language from your IDE, terminal, or any MCP-compatible tool. It reads Sezzle's documentation only; it never calls the Sezzle API.

### Setup

<Steps>
  <Step title="Install the MCP server">
    Run the following command. It detects the AI tools configured in the current project (Cursor, Claude Code, VS Code, and others) and asks which ones to install to. Add `-g` to install globally instead of into the current project.

    ```bash theme={"system"}
    npx add-mcp https://docs.sezzle.com/mcp --name sezzle-docs
    ```

    To configure a client by hand instead:

    <Tabs>
      <Tab title="Claude Code">
        Add the server:

        ```bash theme={"system"}
        claude mcp add --transport http sezzle-docs https://docs.sezzle.com/mcp
        ```

        Then confirm it connected:

        ```bash theme={"system"}
        claude mcp list
        ```
      </Tab>

      <Tab title="Cursor">
        Press <kbd>Command</kbd> + <kbd>Shift</kbd> + <kbd>P</kbd> (<kbd>Ctrl</kbd> + <kbd>Shift</kbd> + <kbd>P</kbd> on Windows), search for `Open MCP settings`, and click **Add custom MCP** to open `mcp.json`. Add:

        ```json theme={"system"}
        {
          "mcpServers": {
            "sezzle-docs": {
              "url": "https://docs.sezzle.com/mcp"
            }
          }
        }
        ```
      </Tab>

      <Tab title="VS Code">
        Create a `.vscode/mcp.json` file containing:

        ```json theme={"system"}
        {
          "servers": {
            "sezzle-docs": {
              "type": "http",
              "url": "https://docs.sezzle.com/mcp"
            }
          }
        }
        ```
      </Tab>
    </Tabs>
  </Step>

  <Step title="Begin prompting">
    Ask a question, and your agent searches Sezzle's documentation for the answer.

    **Example prompts:**

    * "How do I create a Sezzle checkout session, and what do I do next?"
    * "How do I capture an authorized order?"
    * "What's the difference between Authorize Only and Authorize and Capture?"
    * "How do I integrate the Sezzle iOS SDK?"
    * "Which webhook events does Sezzle send, and how do I subscribe to them?"
    * "How do I add the Sezzle widget to a WooCommerce store?"
  </Step>
</Steps>

## Skills

A skill gives your AI agent a procedure to follow for a specific task: the API calls in order, the decisions along the way, and the steps you have to complete yourself.

<Tip>
  The MCP server lets your agent *search* our docs. A skill tells it *what to do*.
</Tip>

### Install skills

Run the following command:

```bash theme={"system"}
npx skills add https://docs.sezzle.com
```

<Note>
  If your AI agent runs the command, pass `--agent` with its name — it can't answer the installer's prompt, so it installs to every agent.
</Note>

Common values are `claude-code`, `cursor`, `github-copilot`, `codex` and `gemini-cli`. The [`skills` CLI reference](https://github.com/vercel-labs/skills) has the full list.

* `--agent claude-code cursor` installs to more than one agent.
* `--skill sezzle-prototype-sandbox` installs one skill instead of all of them.
* `-g` installs globally instead of into the current project.

To add a skill manually, copy the `skill.md` file linked below and paste it into your agent.

### Available skills

We also publish this list as JSON at [`docs.sezzle.com/.well-known/skills/index.json`](https://docs.sezzle.com/.well-known/skills/index.json).

| Name                                                                                                       | Description                                                                                                                                                                                                                                                                                  |
| ---------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| [`sezzle-prototype-sandbox`](https://docs.sezzle.com/.well-known/skills/sezzle-prototype-sandbox/skill.md) | Get a working Sezzle checkout in sandbox from a standing start: sign up, generate API keys, create a session, complete checkout with test data, then release, capture, and refund the order. Confirms your sandbox account is ready to build against, and tells you what to fix if it isn't. |
| [`sezzle-checkout`](https://docs.sezzle.com/.well-known/skills/sezzle-checkout/skill.md)                   | Implement a full Sezzle BNPL checkout integration: authentication, session creation, shopper redirect, and order management (capture, refund, release, reauthorize, upcharge). Covers direct REST API and the JavaScript Checkout SDK (popup, iframe, redirect), plus express checkout.      |
| [`sezzle-webhooks`](https://docs.sezzle.com/.well-known/skills/sezzle-webhooks/skill.md)                   | Set up Sezzle webhooks for order, tokenization, and dispute events, with HMAC-SHA256 signature verification and delivery troubleshooting.                                                                                                                                                    |

## Markdown and llms.txt

Append `.md` to any page URL to get its markdown source. Paste that straight into an LLM.

We also support the [/llms.txt](https://llmstxt.org) standard.

<CardGroup cols="1">
  <Card title="/llms.txt" icon="map" horizontal>
    An index of every page in our docs, hosted at [https://docs.sezzle.com/llms.txt](https://docs.sezzle.com/llms.txt).
  </Card>

  <Card title="/llms-full.txt" icon="clipboard-list" horizontal>
    The full text of our docs in one file, hosted at [https://docs.sezzle.com/llms-full.txt](https://docs.sezzle.com/llms-full.txt).
  </Card>
</CardGroup>
