# Other MCP-Compatible Clients

Any client that implements the [<mark style="color:blue;">Model Context Protocol</mark>](https://modelcontextprotocol.io/) specification can connect to the MCP server.

## Generic configuration

The server communicates over **stdio** using the MCP JSON-RPC protocol. Configure your client to start the binary and pass your credentials as environment variables:

| Parameter             | Value                                                                               |
| --------------------- | ----------------------------------------------------------------------------------- |
| Transport             | `stdio` (standard input/output: the server runs as a local process on your machine) |
| Command               | `/path/to/ionoscloud-mcp`                                                           |
| `IONOS_TOKEN`         | Your IONOS CLOUD API token (required)                                               |
| `IONOS_S3_ACCESS_KEY` | Your Object Storage access key (optional)                                           |
| `IONOS_S3_SECRET_KEY` | Your Object Storage secret key (optional)                                           |

## JSON format

Most MCP clients use a JSON configuration similar to:

```json
{
  "mcpServers": {
    "ionoscloud": {
      "command": "/path/to/ionoscloud-mcp",
      "env": {
        "IONOS_TOKEN": "your-api-token"
      }
    }
  }
}
```

Refer to your client's documentation for the exact configuration format and file location.

For programmatic use via the Anthropic SDK or other LLM APIs, see the [<mark style="color:blue;">MCP SDK documentation</mark>](https://github.com/modelcontextprotocol/python-sdk) for how to connect to a stdio-based MCP server as a subprocess.

## Tool discovery

The server exposes all tools through the standard MCP `tools/list` method. By default, all tools (including Compute Engine and Object Storage) are registered at startup. If your client has a tool-count cap or you want to defer Compute and Object Storage registration, set `IONOS_MCP_LOAD_MODE=lazy`; the client must support `notifications/tools/list_changed` to discover the additional tools after a loader-sentinel call. For more information, see [<mark style="color:blue;">Selective Tool Loading</mark>](/cloud/ai/mcp-server/configuration/selective-tool-loading.md).

## Clients with limitations

**OpenAI Codex CLI:** supports MCP via a `mcpServers` block in `~/.codex/config.toml` (TOML format, not JSON). Stdio transport supported. GA.

**ChatGPT:** supports MCP only for **remote HTTP/SSE servers**, not local stdio binaries. The MCP server runs locally over stdio; you cannot connect it directly to ChatGPT without a public HTTPS endpoint. Not supported for typical developer setups.

**n8n:** can call stdio MCP servers as an MCP client from within an n8n workflow using the MCP Client node. However, the n8n MCP Server Trigger (for exposing n8n as an MCP server) does not support stdio. If your workflow needs to call the MCP server, configure the MCP Client node with the binary path and credentials.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.ionos.com/cloud/ai/mcp-server/connect-to-an-ai-client/other-clients.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.