# Use the docs with AI tools

Connect your AI coding assistant to these docs so it can search and read them while it helps you integrate. Three ways, from the most to the least capable: the docs MCP server, `llms.txt`, and a Markdown copy of every page. All three cover only the public docs and need no account or key.

## Docs MCP server
The docs MCP server lets an assistant that supports the Model Context Protocol (MCP) look things up in these docs on its own. Address:

```text
https://docs.wandura.ai/mcp
```

It uses the Streamable HTTP transport, needs no sign-in, and is read-only: it cannot see your account, your keys or your films.

### Claude Code

Run once in a terminal:

```bash
claude mcp add --transport http wandura-docs https://docs.wandura.ai/mcp
```

Add `--scope project` to share it with your team through the project's `.mcp.json`.

### Cursor

Add this to `~/.cursor/mcp.json` (all projects) or `.cursor/mcp.json` (one project):

```json
{
  "mcpServers": {
    "wandura-docs": {
      "url": "https://docs.wandura.ai/mcp"
    }
  }
}
```

### VS Code

Add this to `.vscode/mcp.json`:

```json
{
  "servers": {
    "wandura-docs": {
      "type": "http",
      "url": "https://docs.wandura.ai/mcp"
    }
  }
}
```

### Claude Desktop

Open **Settings → Connectors → Add custom connector**, name it `Wandura Docs` and paste `https://docs.wandura.ai/mcp` as the URL.

### Tools

| Tool | What it does |
|---|---|
| `search_docs` | Searches the docs and returns the best-matching sections with a link and a short excerpt. Options: `lang` (`en` or `vi`, default `en`), `limit` (1–10, default 5), `version` (an older API version, if you need one). |
| `read_page` | Returns one page as Markdown, for example `/developers/quickstart` or `/vi/developers/quickstart`. |
| `submit_feedback` | Sends feedback about a page to the Wandura team: `kind` is `bug` (something is wrong or unclear) or `feature` (a suggestion). It works like the **Send feedback** button and shares its limit of a few messages every 10 minutes. |

Try it by asking your assistant: *"Using wandura-docs, how do I request playback for an episode?"*

> **Note:** Only send feedback through your assistant when you mean to. Never put API keys, tokens or personal data in feedback.

## llms.txt
For tools without MCP support, point them at these plain-text files:

- `https://docs.wandura.ai/llms.txt`: a list of every page with a one-line summary.
- `https://docs.wandura.ai/llms-full.txt`: every page in one file.
- Vietnamese: `https://docs.wandura.ai/vi/llms.txt` and `https://docs.wandura.ai/vi/llms-full.txt`.

## Markdown copy of each page
Add `.md` to the address of any page to get it as Markdown, for example `https://docs.wandura.ai/developers/quickstart.md`. The **Copy page** button at the top of each page copies the same text, ready to paste into a chat.