AI is now part of everyday development and is here to stay. This session will demonstrate how to utilize Claude Code, Claude Desktop, and ChatGPT as practical co-pilots for real-world Drupal work, with a focus on Drupal MCP, the Tool API module, and the Audit Export module.
We will start with easy, low-friction workflows you can use right away. You will see how to use LLMs to debug custom modules, refactor legacy code, reason about configuration, and write tests against your existing Drupal projects. Along the way, we will compare Claude and ChatGPT at a high level, discuss their strengths and trade-offs, and explain why many developers currently lean toward Claude for in-depth coding and Drupal development while still using ChatGPT for quick ideas, documentation, and non-code help.
We will also look at how to choose between Claude Code and Claude Desktop. Both can handle multi-file workflows, so we will focus on context length and task size. You will see when larger, more complex coding tasks are a better fit for Claude Code, and when smaller, targeted changes and focused prompts work better in Claude Desktop.
From there, we will step inside Drupal itself. You will learn about the Tool API module, how tools are defined in Drupal, and how Drupal MCP enables AI clients to call those tools in a safe and structured way. Using the Audit Export module as a concrete example, we will demonstrate how it can expose structured data about your site, enabling AI to perform configuration audits, highlight potential issues, and suggest next steps for your Drupal installations.
You will leave with a clear picture of:
- When to reach for Claude vs ChatGPT in a Drupal context
- How to decide between Claude Code and Claude Desktop for different tasks
- How Tool API and Drupal MCP work together to let AI "talk" to Drupal
- How Audit Export can power AI-driven audits of your Drupal sites
No prior AI experience is required, only basic familiarity with Drupal development.
