Hosted Model Context Protocol

Connect to Ryter Pro MCP. Nothing to install.

The hosted Remote MCP server gives compatible AI clients direct access to Ryter Pro text humanization. Add one HTTPS endpoint, provide your API key, and start using the tools.

Current releasev0.1.0

An MCP client with Streamable HTTP and custom-header support, plus a Ryter Pro API key. Python is not required.

Remote MCP endpoint

https://mcp.ryter.pro/mcp

Remote MCP is the recommended setup

Use the hosted server unless your MCP client supports only local stdio processes.

Hosted and ready

No Python environment, package installation, or background process.

Your client connects over HTTPS and authenticates each request with your own Ryter Pro API key.

https://mcp.ryter.pro/mcp

Put your API key in the X-API-Key header. Do not add it to the endpoint URL.

Remote MCP quick start

The connection needs only the hosted URL and one request header.

Create an API key

Generate a Ryter Pro key from your account before connecting.

Add the endpoint

Use the hosted Streamable HTTP URL in your MCP client.

Set X-API-Key

Send your key in the X-API-Key header on every MCP request.

json

{
  "mcpServers": {
    "ryterpro": {
      "type": "http",
      "url": "https://mcp.ryter.pro/mcp",
      "headers": {
        "X-API-Key": "YOUR_API_KEY"
      }
    }
  }
}

Replace YOUR_API_KEY with a key from the Ryter Pro API Keys page. Some clients call the transport streamable-http instead of http.

Connect Claude Code

shell

claude mcp add \
  --transport http \
  --scope user \
  --header "X-API-Key: YOUR_API_KEY" \
  ryterpro https://mcp.ryter.pro/mcp

claude mcp list

The user scope makes Ryter Pro available across your Claude Code projects. Open or restart Claude Code and run /mcp to inspect the connection and its two tools.

Try it in Claude Code

Use the Ryter Pro humanize_text tool to rewrite this paragraph in a natural professional tone while preserving its meaning.

Connect Codex

Add the hosted endpoint to Codex config and source X-API-Key from an environment variable.

Set your key in the environment that starts Codex:

shell

export RYTERPRO_API_KEY="YOUR_API_KEY"

Add this server to ~/.codex/config.toml:

toml

[mcp_servers.ryterpro]
url = "https://mcp.ryter.pro/mcp"
env_http_headers = { "X-API-Key" = "RYTERPRO_API_KEY" }

Restart Codex, then use /mcp or run codex mcp list to verify the connection. The Codex CLI, IDE extension, and desktop app share MCP configuration for the same local Codex host.

Try it in Codex

Use Ryter Pro's humanize_text tool on the selected copy. Keep the meaning and make the wording sound less formulaic.

Available tools

Both remote and local connections expose the same focused tool set.

humanize_text

Humanizes or rewrites text through the Ryter Pro API. Accepts text and an optional mode such as basic, advanced, aggressive, or beta. Successful calls consume credits from the API key owner.

ryterpro_api_info

Reports the active transport, API base URL, key status, and official Ryter Pro documentation links.

Optional: run the MCP server locally

Use the PyPI package only when your client cannot connect to remote HTTP servers or you specifically prefer a local stdio process.

shell

pip install ryterpro-mcp
export RYTERPRO_API_KEY="YOUR_API_KEY"
ryterpro-mcp

json

{
  "mcpServers": {
    "ryterpro": {
      "command": "ryterpro-mcp",
      "env": {
        "RYTERPRO_API_KEY": "YOUR_API_KEY"
      }
    }
  }
}

The local package uses MCP stdio and exposes the same tools as the hosted endpoint. It requires Python 3.10 or later and must remain installed in an environment your MCP client can launch.

Using the tool

Once connected, ask your client to use Ryter Pro for a specific piece of text.

Example request

Use the Ryter Pro humanize_text tool in advanced mode to make this draft sound more natural while preserving its meaning: “Our platform enables users to efficiently improve written content.”

Mention humanize_text explicitly if the client does not select the tool automatically.

Troubleshooting

Start with the remote endpoint and header before checking client-specific settings.

Confirm the server URL is exactly https://mcp.ryter.pro/mcp.
Confirm that the header name is exactly X-API-Key and that the key is active.
Restart the MCP client after adding or changing the server configuration.
Use ryterpro_api_info to confirm the connection can see an API key.
A 401 response means X-API-Key is missing or was not sent by the client.
A 402 response means the Ryter Pro account does not have enough credits.
For the optional local server, confirm Python 3.10+ and the ryterpro-mcp command are available to the client.