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

# Documentation MCP

> A public, no-auth MCP server that gives AI assistants semantic search and read access to the lobstr.io API documentation. To run scrapers in your account, see the MCP Server page instead.

This server answers *"how does lobstr.io work?"* — it exposes the API
documentation to AI assistants for semantic search and reading. It is read-only
and unauthenticated; it cannot touch your account.

<Note>
  Looking to **run scrapers** from Claude or ChatGPT — discover a scraper,
  configure it, run it, and pull results in the conversation? That's the
  operational [MCP Server](/docs/mcp), which connects to your account over OAuth.
  This page is only the documentation assistant.
</Note>

**Endpoint:** `https://docs.lobstr.io/mcp`

<CardGroup cols={3}>
  <Card title="2 Tools" icon="wrench">
    Semantic search and full filesystem access to every docs page.
  </Card>

  <Card title="No Auth" icon="unlock">
    Public endpoint — no API key needed to query the docs.
  </Card>

  <Card title="SSE Transport" icon="globe">
    HTTP + Server-Sent Events transport, compatible with all MCP clients.
  </Card>
</CardGroup>

## Setup

<CodeGroup>
  ```bash Claude Code theme={null}
  claude mcp add lobstrio-docs \
    --transport http \
    https://docs.lobstr.io/mcp
  ```

  ```json Claude Desktop theme={null}
  {
    "mcpServers": {
      "lobstrio-docs": {
        "type": "http",
        "url": "https://docs.lobstr.io/mcp"
      }
    }
  }
  ```

  ```json Cursor (.cursor/mcp.json) theme={null}
  {
    "mcpServers": {
      "lobstrio-docs": {
        "type": "http",
        "url": "https://docs.lobstr.io/mcp"
      }
    }
  }
  ```

  ```bash Any MCP Client theme={null}
  curl -X POST https://docs.lobstr.io/mcp \
    -H "Content-Type: application/json" \
    -H "Accept: application/json, text/event-stream" \
    -d '{
      "jsonrpc": "2.0",
      "method": "tools/call",
      "params": {
        "name": "search_lobstr_io",
        "arguments": { "query": "create squid" }
      },
      "id": 1
    }'
  ```
</CodeGroup>

## Tools

| Tool                              | Description                                                                                                           |
| --------------------------------- | --------------------------------------------------------------------------------------------------------------------- |
| `search_lobstr_io`                | Semantic search across all documentation — titles, descriptions, guides, and API references.                          |
| `query_docs_filesystem_lobstr_io` | Read-only shell-like access to every docs page as `.mdx` files. Supports `cat`, `head`, `rg`, `tree`, `ls`, and more. |

## Example Prompts

Once connected, ask your AI assistant:

* *"How do I create a squid and start scraping Google Maps?"*
* *"Show me the Python code for adding tasks to a scraper"*
* *"What parameters does the LinkedIn Profile Scraper accept?"*
* *"List all delivery configuration options (email, S3, webhook)"*
* *"How do I handle rate limiting in the lobstr.io API?"*

## JSON-RPC Methods

| Method       | Description                                      |
| ------------ | ------------------------------------------------ |
| `initialize` | Handshake — returns server info and capabilities |
| `tools/list` | List all available tools with input schemas      |
| `tools/call` | Execute a tool with arguments                    |