James Abrahams Shares Drupal AI Roadmap and No-Code Agent Training Update

James Abrahams Shares Drupal AI Roadmap and No-Code Agent Training Update

James Abrahams, this week, shared a detailed progress report on the key initiatives underway in the Drupal AI project, covering product‑facing experiments, core framework development, and advanced AI agent tooling aimed at evolving how AI integrates with Drupal CMS.

The update opens with a structured roadmap outlining the next phases for Drupal AI, beginning with user-facing features that enhance functionality and usability. Among them is a triaged “XB AI Issue List”, the AI Context 1.0 roadmap for Drupal CMS 2.0, and a new admin UI that allows enabling or disabling individual AI providers or improved installation and configuration workflows. A refined AI Recipe installer and a proof-of-concept for an AI-powered design system are also in development.

Technical progress includes the ongoing creation of a Tool API moving toward beta release, relocation and enhancement of AI Agents into the AI module core, and initial steps toward supporting autonomous background agents via Symfony Messenger. Plans to abstract AI logic into Symfony AI and eventually offer a Symfony AI Agents SDK reflect a broader architectural transition. Innovations also include a browser-based LLM provider with security controls and dedicated Drupal CMS LLM fine-tuning, with testing tools being simplified for easier agent development and reuse.

Long-term proposals aim for eventual core inclusion. The most prominent is “Workspaces for Config”, which envisions a Git-like versioning experience for all Drupal configuration. Also in the pipeline are real-time collaborative editing features, an enhanced admin design system, and improvements to core moderation and workflow tooling.

The update also includes a slate of innovation-focused deliverables, developed primarily for demonstration but with potential for productisation. Highlights include a Mass Alt Text Evaluator, a Content Classification Agent, and experimental work on a Site Merging Agent. These tools are built on the MCP and Tools API stack, providing accessible automation options for accessibility, metadata, and content management. A Content Editor Agent using Workspaces ensures that edits made by AI agents are reviewed before publication, adding a safety net for automated content handling.

To help the community explore this technology hands-on, a livestream training session titled “No Code AI Agent Training” is scheduled for 13 January 2026 at 16:00 Brussels time. The session will walk users through the process of building AI agents without writing code, using tools like Flowdrop UI, MCP, and guardrails, with a focus on the Alt Text Evaluator use case. It will be streamed on YouTube with further content planned for SparkFabrik and OpenSense Labs channels.

“We first show how to build the AI Agents yourself using flowdrop UI, standard agents, Tools API, MCP and maybe guardrails.”

– James Abrahams

By aligning product development, framework architecture, and experimental tooling, the Drupal AI team is laying the foundation for a flexible, user-friendly AI experience within Drupal. These efforts reflect a strategic direction aimed at making AI tools both accessible to non-developers and powerful enough for deep technical integration.

For more details, visit the AI Initiative Issue Tracker on Drupal.org.

Original update by James Abrahams via LinkedIn. 

Disclosure: This content is produced with the assistance of AI.

Note: The vision of this web portal is to help promote news and stories around the Drupal community and promote and celebrate the people and organizations in the community. We strive to create and distribute our content based on these content policy. If you see any omission/variation on this please reach out to us at #thedroptimes channel on Drupal Slack and we will try to address the issue as best we can.

Related Drupal Initiatives

Related Organizations

Upcoming Events

Latest Opportunities