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

> ## Agent Instructions
> Marsell's MCP server is at https://getmarsell.com/mcp: agents connect with a key the owner makes in the app (a project's Agents page, Your own AI). See /developers/mcp for its tools. There is no public REST API yet; don't suggest REST endpoints.
> The search MCP server at https://docs.getmarsell.com/mcp only searches these docs. It can't read or change anyone's Marsell projects.
> Each business is a project with seven places: Home, Create, Market, Sell, People, Agents and Brand. 'Marsell plan' means the subscription (Solo, Core, Growth, Network); 'the plan' or 'the week' means the work Marsell has planned.
> Tokens are Marsell's unit for AI and media work, granted monthly by the Marsell plan. They are not model tokens.

# Use these docs with AI

> Read these docs as Markdown, point an AI tool at llms.txt, or connect the docs' search MCP server to Claude, ChatGPT, Cursor or VS Code.

Every page on this site is written for people and for AI tools. You can hand a page to an assistant, give it the whole site as one file, or connect the docs' search server so it looks things up while it answers.

<Note>
  Everything on this page reads these docs only. None of it can see or change your Marsell projects: for that, connect the [Marsell MCP server](/developers/mcp).
</Note>

## Send a page to an assistant

The menu at the top of each page copies the page as Markdown, opens the Markdown version, or opens the page in ChatGPT, Claude or Perplexity with the page attached.

## Read any page as Markdown

Add `.md` to the end of a page's address, for example [docs.getmarsell.com/quickstart.md](https://docs.getmarsell.com/quickstart.md). Tools can also ask for Markdown with an `Accept: text/markdown` header:

```bash theme={null}
curl -L -H "Accept: text/markdown" https://docs.getmarsell.com/quickstart
```

## Give a tool the whole site

* [`llms.txt`](https://docs.getmarsell.com/llms.txt) lists every page with a one-line description, in the order of the sidebar.
* [`llms-full.txt`](https://docs.getmarsell.com/llms-full.txt) holds the full text of every page in one file.

## Connect the search MCP server

The docs' search server is at `https://docs.getmarsell.com/mcp`. It needs no key. Once connected, your assistant can search these docs and read whole pages while it answers.

<Tabs>
  <Tab title="Claude">
    <Steps>
      <Step title="Add a custom connector">
        Open [Settings > Connectors](https://claude.ai/settings/connectors) in Claude and click **Add custom connector**.
      </Step>

      <Step title="Enter the server">
        Name it "Marsell docs", paste `https://docs.getmarsell.com/mcp` as the URL and click **Add**.
      </Step>

      <Step title="Use it in a chat">
        In a chat, click the plus button and turn on **Marsell docs**.
      </Step>
    </Steps>
  </Tab>

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

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

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

  <Tab title="Cursor">
    Open the command palette, run **Open MCP settings**, click **Add custom MCP** and add:

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

  <Tab title="VS Code">
    Create `.vscode/mcp.json` in your project:

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

Other assistants that support remote MCP servers over HTTP take the same address.


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