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

# Deck Docs MCP server

> Search the Deck docs from Claude, Cursor, and other AI tools over the Model Context Protocol.

The Deck docs are available as a hosted [MCP](https://modelcontextprotocol.io) server:

```text theme={null}
https://docs.deck.co/mcp
```

Connect it to any MCP client and your AI tool can search this site and read full pages while it works, instead of relying on training data or web search. The server is read-only and covers documentation only. It has no access to your Deck account or your data, and it cannot call the Deck API.

## Connect it

You can also connect from any docs page: open the context menu at the top of the page to copy the server URL or add it to Cursor or VS Code in one click.

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

  <Tab title="Claude">
    1. Open [Connectors](https://claude.ai/settings/connectors) in the Claude settings.
    2. Click **Add custom connector**.
    3. Name it `Deck Docs` and enter `https://docs.deck.co/mcp` as the URL.
    4. Click **Add**.
  </Tab>

  <Tab title="Cursor">
    1. Open the command palette and search for **Open MCP settings**.
    2. Click **Add custom MCP**. This opens `mcp.json`.
    3. Add the server:

    ```json theme={null}
    {
      "mcpServers": {
        "deck-docs": {
          "url": "https://docs.deck.co/mcp"
        }
      }
    }
    ```
  </Tab>

  <Tab title="Codex">
    Add to `~/.codex/config.toml`:

    ```toml theme={null}
    [mcp_servers.deck-docs]
    url = "https://docs.deck.co/mcp"
    ```
  </Tab>

  <Tab title="Other clients">
    Any client that supports the streamable HTTP transport can connect to `https://docs.deck.co/mcp`. Agents can also discover the server automatically through the standard discovery document at `https://docs.deck.co/.well-known/mcp`.
  </Tab>
</Tabs>

## What it provides

| Tool                              | What it does                                                              |
| --------------------------------- | ------------------------------------------------------------------------- |
| `search_deck_docs`                | Searches the docs and returns matching pages with titles and links        |
| `query_docs_filesystem_deck_docs` | Reads full pages as Markdown, including batch reads across multiple pages |
| `submit_feedback`                 | Reports incorrect, outdated, or confusing documentation to the Deck team  |

The server also exposes the [Deck skill](/ai) as an MCP resource, so connected agents can load Deck integration guidance without installing `skill.md` separately.

Once connected, your tool decides when to search. Ask it things like "How does session reuse work in Deck?" or "Write a webhook handler for `task_run.completed`" and it pulls the answer from the current docs.

## Limitations

* **Docs only.** The server can't see your runs, credentials, or metrics. For account questions in chat form, use [Deck in Slack](/integrations/slack). For programmatic access, use the [API](/api/using-the-api).
* **Read-only.** Apart from `submit_feedback`, the server only searches and reads documentation.

For other ways to give AI tools Deck context, including the Deck skill, a context prompt, and LLM-formatted doc exports, see [Build with AI](/ai).