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

# Docs MCP server

> Connect Claude, Claude Code, Cursor, or any MCP client to this documentation so it can search and read the Pipecorn API docs while it writes your integration.

This documentation runs its own [MCP](https://modelcontextprotocol.io)
server. Connect it to your assistant or coding agent and it can search these
docs and read any page or OpenAPI spec on demand, instead of you pasting
pages into the chat or loading all of [`llms-full.txt`](/api-reference/llms)
into its context.

| | |
| - | - |
| **Server URL** | `https://docs.pipecorn.com/mcp` |
| **Transport** | Streamable HTTP |
| **Authentication** | None. The docs are public, so no API key or sign-in is needed |
| **Cost** | Free. It reads the docs only and never touches your Pipecorn account or credits |

<Note>
  This is not the [Pipecorn MCP server](/mcp/overview). The docs server lets a
  model **read about** the API. The Pipecorn MCP server at
  `https://mcp.pipecorn.com/mcp` lets it **use** Pipecorn: search companies,
  extract leads, enrich contacts. You can connect both.
</Note>

## Tools

| Tool | What it does |
| - | - |
| `search_pipecorn_api_documentation` | Searches every page and returns the matching sections, each with its title and a link |
| `query_docs_filesystem_pipecorn_api_documentation` | Reads the docs as a read-only file tree: `cat` or `head` a page, `rg` for an exact term, `tree` or `ls` to browse. A page at `/essentials/idempotency` is the file `/essentials/idempotency.mdx`, and the OpenAPI specs are there too |
| `submit_feedback` | Reports a page that is wrong, outdated or unclear to the Pipecorn docs team |

## Connect your client

<Tabs>
  <Tab title="Claude">
    In the Claude sidebar, click **Customize**, then go to **Connectors**.
    Click **+**, then **Add custom connector**, set **Name** to
    `Pipecorn Docs` and **URL** to `https://docs.pipecorn.com/mcp`, and save.
    There is nothing to sign in to.
  </Tab>

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

    Add `--scope user` to make it available in every project, and check the
    connection with `/mcp`.
  </Tab>

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

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

  <Tab title="Other clients">
    Point any MCP client that supports Streamable HTTP at
    `https://docs.pipecorn.com/mcp`. No headers are needed.
  </Tab>
</Tabs>

The **Connect to MCP** option in the menu at the top of every page copies the
same URL.

## Try it

Once connected, ask your agent something that needs the docs:

```text theme={null}
Using the Pipecorn docs, write a Python script that enriches each contact in
leads.csv with the async single enrichment endpoint, polls the status URL with
backoff, and retries safely with an Idempotency-Key.
```

The agent searches the docs for enrichment, async results and idempotency,
reads the pages it needs, and writes the script from the current reference.


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