Configure the OpenEdge AI Assistant through CLI

Save PDF

Last Updated: May 14, 2026
2 minute read

OpenEdge
Version 12.8
Documentation

The Command Line Interface (CLI)-based configuration provides a lightweight, script-driven alternative to the Integrated Development Environment-based configuration for the OpenEdge AI Assistant.

Prerequisites

Before you begin, you must meet the following prerequisites:

You have installed Python version 3.9 or later. During installation, you have selected the Add python.exe to PATH checkbox.
You have downloaded the following Python scripts from the Customer Validation Program (CVP) page:
- mcp-login.py
- mcp-proxy.py

To configure the OpenEdge AI Assistant through CLI, perform the following steps:

Open your CLI and navigate to the directory that contains the following Python scripts:
- mcp-login.py
- mcp-proxy.py
Run the login script.
python ./mcp-login.py

Enter the API key in the API key field when prompted. For information on obtaining an API key, see Obtain Progress Data Cloud API key.

For example:


Enter your Progress Data Cloud API Key:
API Key: Y9Bub5PQJuznN4_1xYZABc==

After the authentication is successful, an output similar to the following appears:

🔐 Authenticating with Progress Data Cloud...
🔍 Validating API Key status...


════════════════════════════════════════════════════════════════════════════════
✅ AUTHENTICATION SUCCESSFUL
════════════════════════════════════════════════════════════════════════════════

Create or update the mcp.json file:
- If you do not have an mcp.json file, create one in your configuration directory.
- If you already have an mcp.json file, add the configuration shown in Step 6 to it.
Open your mcp.json file in a text editor.

To enable automatic bearer token refresh for active MCP sessions, add the following configuration and replace the <PATH> placeholder in the args field with the absolute path to the mcp-proxy.py script on your system.


    
"mcpServers": {
    "openedge-abl": {
      "command": "python",
      "args": ["<PATH>/mcp-proxy.py"]
    },
  }
...

For example:


    
"mcpServers": {
    "openedge-abl": {
      "command": "python",
      "args": ["/home/user/scripts/mcp-proxy.py"]
    },
  }

Start the OpenEdge AI Assistant from your IDE. In Windsurf or VS Code, the MCP server starts automatically when the IDE reads the mcp.json configuration file. You can also trigger it manually by reloading the window (CTRL+SHIFT+P → Reload Window).

After successfully starting the OpenEdge AI Assistant, an output similar to the following appears:


    
026-01-05 19:15:48.223 [info] Starting server openedge-abl
2026-01-05 19:15:48.223 [info] Connection state: Starting
2026-01-05 19:15:48.224 [info] Starting server from LocalProcess extension host
2026-01-05 19:15:48.259 [info] Connection state: Starting
2026-01-05 19:15:48.259 [info] Connection state: Running
2026-01-05 19:15:48.375 [warning] [server stderr] [PROXY INFO] API Key loaded from config file
2026-01-05 19:15:48.375 [warning] [server stderr] [PROXY INFO] API Key found (length=24)
2026-01-05 19:15:48.375 [warning] [server stderr] [PROXY INFO] Fetching initial token...
2026-01-05 19:15:49.116 [warning] [server stderr] [PROXY INFO] Token fetched and saved to config (expires in 55 minutes)
2026-01-05 19:15:49.116 [warning] [server stderr] [PROXY INFO] Initial token fetched successfully
2026-01-05 19:15:49.116 [warning] [server stderr] [PROXY INFO] Token refresh worker started
2026-01-05 19:15:49.117 [warning] [server stderr] [PROXY INFO] Token validation worker started
2026-01-05 19:15:49.117 [warning] [server stderr] [PROXY INFO] OpenEdge MCP PDC Proxy started
2026-01-05 19:15:49.117 [warning] [server stderr] [PROXY INFO] MCP Server: https://openedgemcp.marklogic.cloud/mcpinstance/mcp
2026-01-05 19:15:49.117 [warning] [server stderr] [PROXY INFO] Token refresh: Every 50 minutes
2026-01-05 19:15:49.117 [warning] [server stderr] [PROXY INFO] Token validation: Every 3 minutes
...
2026-01-05 19:15:51.754 [info] Discovered 11 tools
...

Note: After configuring the OpenEdge AI Assistant:

If you restart your system, restart the OpenEdge AI Assistant.
If you generate a new API key, you must log in again and restart the OpenEdge AI Assistant.

Use the OpenEdge AI Coding Assistant

Configure the OpenEdge AI Assistant through CLI

Table of Contents

Configure the OpenEdge AI Assistant through CLI