Codify

Codify MCP is a server that lets AI assistants run browser automation tasks via Apify Agent. It accepts tools as clean JSON payloads and executes them without manual setup, enabling you to build reusable browser automation tools that AI can invoke directly.

How to use

You use Codify MCP by running it as a local command or connecting it to your AI assistant. Create or export tools as JSON objects, then invoke Codify MCP with those tool definitions. Tools are passed one per argument and are executed in sequence or individually depending on how you configure your client.

How to install

Prerequisites you need to have installed on your system before using Codify MCP include Node.js and npm. Follow these steps to get started.

# Install the MCP server globally
npm install -g codify-mcp

# Or run directly with npx without installing globally
npx codify-mcp {{JSON_MCP_TOOL_1}} {{JSON_MCP_TOOL_2}} {{JSON_MCP_TOOL_3}}...

Additional setup you might perform

1. Optional Apify token management: you can log in to Apify to save your token locally, or supply it via environment variables. 2) Prepare a tool you will export as JSON and pass to the MCP server as a single argument per tool. 3) If you plan to connect with AI assistants like Claude Desktop, configure the MCP server entry in the assistant’s settings as shown in the example.

Examples you can try

{
  "name": "scrape_product",
  "description": "Scrape product info from a page",
  "inputSchema": {
    "type": "object",
    "properties": {
      "url": {
        "type": "string",
        "description": "Product page URL"
      }
    },
    "required": ["url"]
  },
  "implementation": {
    "type": "apify-actor",
    "actorId": "cyberfly/apify-agent",
    "script": "await page.goto(inputs.url); const title = await page.textContent('h1'); return {title};"
  }
}

How to connect to an AI client

You can wire Codify MCP to an AI client following the example configuration. The MCP server can be invoked via a local command and you pass a tool definition as a JSON string in the args array. Ensure you restart the AI client after you add the MCP entry so the tools are available.

{
  "mcpServers": {
    "codify-mcp": {
      "command": "npx",
      "args": [
        "codify-mcp",
        "{\"name\":\"scrape_product\",\"description\":\"Scrape product info\",\"inputSchema\":{\"type\":\"object\",\"properties\":{\"url\":{\"type\":\"string\",\"description\":\"Product page URL\"}}},\"implementation\":{\"type\":\"apify-actor\",\"actorId\":\"cyberfly/apify-agent\",\"script\":\"await page.goto(inputs.url); const title = await page.textContent('h1'); return {title};\"}}}"
      ],
      "env": {
        "APIFY_TOKEN": "your_token_or_leave_empty_to_use_auth_file"
      }
    }
  }
}

Available tools

scrape_product

Scrape product info from a page by providing a URL and returning product title.