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

> ## Agent Instructions
> To act on Atlan objects, use the Atlan MCP server at https://api.atlan.com/mcp or the atlanai CLI; `atlanai --map json` prints its command map. Run a read-only identity check before any write.
> The docs MCP server at /mcp searches these docs only. It cannot read or change Atlan objects.
> SDK packages: Python `atlanai` (PyPI) and TypeScript `@atlanai/sdk` (npm). Show Python first, then TypeScript.

# For AI agents

> Give coding agents and LLMs clean access to these docs: llms.txt, per-page Markdown, the docs MCP server, the docs skill, and copy-paste prompts.

Every page in these docs is available as Markdown, and the whole site can be
searched from an MCP client. Point your coding agent here before it writes code against Atlan.

<CardGroup cols={2}>
  <Card title="Copy a page" icon="copy" href="#keyboard-shortcuts">
    Copy any page as Markdown from the page menu or with a keyboard shortcut.
  </Card>

  <Card title="Load the docs index" icon="list" href="#endpoints">
    Give an agent `llms.txt` or `llms-full.txt` to load the docs in one request.
  </Card>

  <Card title="Connect the docs MCP" icon="plug" href="#connect-the-docs-mcp-server">
    Let Claude Code, Cursor, or VS Code search these docs on demand.
  </Card>

  <Card title="Use the Atlan MCP" icon="server" href="/mcp/overview">
    Act on Agent Registry objects from an MCP client with Atlan credentials.
  </Card>
</CardGroup>

## Endpoints

These paths are generated from the published docs at `https://platform.atlan.com`. Append
them to that URL.

| Path                          | What it returns                                      | Notes                                                                             |
| ----------------------------- | ---------------------------------------------------- | --------------------------------------------------------------------------------- |
| `/llms.txt`                   | An index of every page, with titles and descriptions | Start here to let an agent choose what to read.                                   |
| `/llms-full.txt`              | The full text of every page in one file              | Large. Use it when the agent can hold the whole site in context.                  |
| `<page>.md`                   | One page as Markdown                                 | Add `.md` to any page URL, for example `/get-started.md`.                         |
| `Accept: text/markdown`       | One page as Markdown                                 | Send this header to any page URL.                                                 |
| `/skill.md`                   | A skill file that describes these docs to an agent   | Install it with the command in [Install the docs skill](#install-the-docs-skill). |
| `/mcp` and `/.well-known/mcp` | The docs MCP server and its MCP discovery document   | No sign-in needed.                                                                |

Request a page as Markdown with the `Accept` header:

```bash theme={null}
curl -L -H "Accept: text/markdown" https://platform.atlan.com/get-started
```

## Connect the docs MCP server

The docs MCP server exposes a search tool over these docs. Connect it once and
your agent can look up the current docs instead of relying on training data.

<Tabs>
  <Tab title="Claude Code">
    ```bash theme={null}
    claude mcp add --transport http atlan-docs https://platform.atlan.com/mcp
    ```
  </Tab>

  <Tab title="Cursor">
    Add the server to `.cursor/mcp.json` in your project, or to
    `~/.cursor/mcp.json` for every project:

    ```json .cursor/mcp.json theme={null}
    {
      "mcpServers": {
        "atlan-docs": {
          "url": "https://platform.atlan.com/mcp"
        }
      }
    }
    ```

    Or choose **Connect to Cursor** from the page menu at the top of any page.
  </Tab>

  <Tab title="VS Code">
    Add the server to `.vscode/mcp.json` in your workspace:

    ```json .vscode/mcp.json theme={null}
    {
      "servers": {
        "atlan-docs": {
          "type": "http",
          "url": "https://platform.atlan.com/mcp"
        }
      }
    }
    ```

    Or choose **Connect to VS Code** from the page menu at the top of any page.
  </Tab>

  <Tab title="Other clients">
    Point any client that supports remote MCP servers over HTTP at
    `https://platform.atlan.com/mcp`. The page menu also has **Copy MCP Server** and
    **Copy MCP install command**.
  </Tab>
</Tabs>

<Warning>
  The docs MCP server only searches documentation. The Atlan product MCP server
  at `https://api.atlan.com/mcp` acts on Agent Registry objects and needs Atlan
  credentials. See [Atlan MCP](/mcp/overview) to connect it.
</Warning>

## Install the docs skill

Install the docs skill to give your agent a short description of these docs and
how to use them:

```bash theme={null}
npx skills@1.7.0 add https://platform.atlan.com
```

## Copy-paste prompts

Paste one of these prompts into your coding agent. None of them need
credentials in the prompt. Keep `ATLAN_API_KEY` and `ATLAN_WORKSPACE_ID` in
your environment.

<Prompt description="Load the docs index, then plan the work" actions={["copy", "cursor"]}>
  Read [https://platform.atlan.com/llms.txt](https://platform.atlan.com/llms.txt), then pick the pages that cover my task and read
  them as Markdown by adding .md to each page URL. Summarize what you will build
  and which Atlan docs pages you relied on before you write any code.
</Prompt>

<Prompt description="Read one page and answer a question" actions={["copy", "cursor"]}>
  Read [https://platform.atlan.com/get-started.md](https://platform.atlan.com/get-started.md) and answer my question using only that page.
  If the page does not cover it, say so and name the page I should read next.
  My question:
</Prompt>

<Prompt description="Instrument an agent with the Atlan SDK" actions={["copy", "cursor"]}>
  Read [https://platform.atlan.com/sdk/overview.md](https://platform.atlan.com/sdk/overview.md), [https://platform.atlan.com/sdk/quickstart.md](https://platform.atlan.com/sdk/quickstart.md), and
  [https://platform.atlan.com/sdk/tracing.md](https://platform.atlan.com/sdk/tracing.md). Then add Atlan tracing to my agent with the Atlan
  SDK: Python package atlanai and TypeScript package @atlanai/sdk. Show the
  Python version first, then TypeScript. Read ATLAN\_API\_KEY and
  ATLAN\_WORKSPACE\_ID from the environment and never hard-code them.
</Prompt>

## Keyboard shortcuts

| Action                                               | Mac                         | Windows and Linux            |
| ---------------------------------------------------- | --------------------------- | ---------------------------- |
| Open search                                          | <kbd>Cmd</kbd>+<kbd>K</kbd> | <kbd>Ctrl</kbd>+<kbd>K</kbd> |
| Copy the page as Markdown (when no text is selected) | <kbd>Cmd</kbd>+<kbd>C</kbd> | <kbd>Ctrl</kbd>+<kbd>C</kbd> |

Agents: start with `/llms.txt` to find the right page, then read it as
Markdown. Use the Atlan MCP server, not the docs MCP server, to act on Atlan objects.