Dajet

Download the latest release from GitHub Releases, create a working directory, extract the distribution, and run the executable. The server listens on HTTP (default port 3000).

# Download from GitHub Releases:
# https://github.com/zhichkin/dajet-mcp-server/releases/latest

# Create working directory and extract
mkdir C:\dajet-mcp-server
# Extract distribution into this directory

# Start the server
C:\dajet-mcp-server\DaJet.Mcp.Server.exe

Pull the Docker image from Docker Hub and run it as a container, exposing port 5000.

docker pull zhichkin/dajet-mcp-server
docker run --name dajet-mcp-server --user=root -it -p 5000:5000 zhichkin/dajet-mcp-server

Edit appsetting.json in the installation root to set the HTTP host and port. The default is http://localhost:3000. Change this if you need a different port or need to bind to all interfaces.

Edit datasources.json in the installation root to add your 1C:Enterprise database connections. Each entry needs a Name (used by AI agents to reference the database), a Type (SqlServer or PostgreSQL), and a ConnectionString.

{
  "DataSources": [
    {
      "Name": "MS_TEST",
      "Type": "SqlServer",
      "ConnectionString": "Data Source=server;Initial Catalog=database;Integrated Security=True;Encrypt=False;"
    }
  ]
}

Add DaJet MCP Server to your MCP client configuration using its HTTP endpoint. In Cursor use .cursor/mcp.json. For other clients use the appropriate config file.

{
  "mcpServers": {
    "dajet-mcp-server": {
      "url": "http://localhost:3000"
    }
  }
}