home / mcp / markup.ai mcp server

Markup.ai MCP Server

Acrolinx NextGen MCP

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "acrolinx-nextgen-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "github:markupai/mcp"
      ],
      "env": {
        "DEBUG": "false",
        "MAX_RETRIES": "3",
        "POLL_INTERVAL": "2000",
        "MAX_TEXT_LENGTH": "100000",
        "MARKUPAI_API_KEY": "your_api_key_here",
        "WORKFLOW_TIMEOUT": "60000",
        "MARKUPAI_BASE_URL": "https://api.markup.ai"
      }
    }
  }
}

The Markup.ai MCP Server lets your IDEs and assistants tap into Markup.ai for automatic text rewriting, quality analysis, and tailored writing suggestions. It supports multiple style guides, dialects, and tones, making it easy to improve text within your existing workflow.

How to use

Use the Markup.ai MCP Server from any MCP-compatible IDE to rewrite, analyze, or receive editing suggestions for your text. You can run it locally or connect to a remote MCP endpoint. Your IDE will prompt you to run Markup.ai analyses, and you’ll approve them with a simple action when needed. You can specify the dialect, tone, and style guide to tailor results to your needs.

How to install

Prerequisites you need before you start are verified and installed on your system.

Node.js 18.0.0 or higher
A Markup.ai API key

Then follow these concrete steps to set up and run the MCP server.

git clone https://github.com/markupai/mcp.git
cd mcp
```

```
npm install
```

```
cp .env.example .env
```
Edit the .env file and add your MARKUPAI_API_KEY.

```
npm run build

Configuration

The MCP server uses environment variables to configure access, timing, and behavior. The following variables are shown as the standard configuration options.

MARKUPAI_API_KEY=your_api_key_here
MARKUPAI_BASE_URL=https://api.markup.ai
DEBUG=false
MAX_TEXT_LENGTH=100000
WORKFLOW_TIMEOUT=60000
POLL_INTERVAL=2000
MAX_RETRIES=3

Troubleshooting and notes

Common issues you might encounter and how to address them.

- If you see a "Calling undefined" message in your IDE, this is normal when Cursor or another MCP client requests tool execution. Click the Run tool button to approve the analysis.

- If you encounter a "Client closed" error, clear the npx cache and restart your IDE.

- Ensure your MARKUPAI_API_KEY is correctly set in the environment. Verify permissions if you encounter API issues.

- For development or build issues, run with debug enabled to gather more logs.

Available tools

markupai_rewrite

Automatically rewrite and improve text content to enhance clarity, tone, and alignment with a chosen style guide.

markupai_check

Analyze text for quality issues without applying changes, returning a detailed quality assessment.

markupai_suggestions

Provide detailed editing suggestions to improve text quality and coherence.

markupai_workflow_status

Check the status of an asynchronous workflow such as rewrites, checks, or suggestions.