The OpenEdge AI Coding Assistant is an AI-powered coding agent that uses large language models (LLMs) to assist or automate tasks in the software development lifecycle.

This guide uses Windsurf as the customized IDE for demonstrating AI-powered code assistance with ABL. However, the underlying setup is designed to be tool-agnostic, allowing you to integrate and benefit from the same AI-driven workflows and rule-based enhancements across a variety of development environments, including Cursor, GitHub Copilot, Amazon Q Developer, Continue (VS Code plugin), and more.

Within Windsurf, the OpenEdge AI Coding Assistant leverages Cascade, which is an agentic framework that understands user intent and codebase context to support:

  • Code generation
  • Error detection and resolution
  • Project planning and task automation

Cascade helps developers stay in flow by handling complex and repetitive coding tasks, allowing more focus on creativity and problem-solving.

To enable seamless collaboration between the OpenEdge AI Coding Assistant and Cascade, the system uses the Model Context Protocol (MCP). MCP acts as a standardized communication bridge between AI agents, RAG systems, and development tools. In practice, developers configure MCP servers in Windsurf using a simple JSON file containing server URLs and authentication tokens.

MCP enables RAG by:
  • Retrieving relevant OpenEdge documentation, rules, and context from a vector database.
  • Augmenting the retrieved context and passing it to LLMs (ChatGPT, Claude, and so on) for accurate, context-aware code generation.
The OpenEdge AI Coding Assistant also supports features like syntax checking, commit message automation, and agentic workflows, making the assistant smarter and more responsive to developer needs.
This approach:
  • Anchors responses in real project data, reducing hallucinations
  • Improves accuracy and relevance of code suggestions
  • Supports onboarding by surfacing documentation automatically
  • Enables smarter automation of coding tasks