fidzu : web

MCP is released in a new 1.2 version

Omedia has put a lot of effort into getting the 1.2 version of MCP out. The module is now security covered, it works with the Tool API, adds OAuth authentication, a much better configuration system and a first preview of MCP Studio.

It now fully supports the HTTP and STDIO transports and independent where you want to source your tools, AI function calling, Drush commands or Tool API its all just available.

Read more about Giorgi Jibladze's post on LinkedIn.

LMStudio is Stable

Thanks to Andrei Ivnitskii (ivnish), the first stable version of the LMStudio provider for the Drupal AI module is now available. This release introduces full integration with LMStudio, giving developers a local, flexible, and GUI-driven environment for running and testing AI models directly from their Drupal site.

This integration is built for developers and data scientists who need a controlled testing environment, but find Ollama or vLLM too complex to set up and use. By pairing Drupal AI with LMStudio, you can iterate quickly, test safely, and work offline - perfect for preparing models before scaling up to cloud-based providers.

Try it out and help out in https://www.drupal.org/project/ai_provider_lmstudio

Meet the QA team!

I met with the awesome AI Initiative QA team a couple of weeks ago and they have set up processes for how both manual and automated QA can happen on issues.

They have an issue where you can follow their work on here: https://www.drupal.org/project/ai_initiative/issues/3550700

The idea is to standardize the "Needs QA" tag, where you as a developer can request manual testing and get feedback on the process.

Next steps, include looking into how DrupalForge/DrupalPod can be used for AI purposes to start images based on an issue on DrupalForge with the push of a button and also local development environments.

If you want to get in touch with the team they are on #ai-quality-insurance Slack channel.

Work on the 2.0 release is on its way

We have already started working on deprecations of modules, and other code for an 1.3.0 release and the work on all the refactors needed for 2.0 is started. The goal is to have a version ready by the end of the year and that fixes up a lot of architectural decisions made in 1.x branch and also removes a lot of features that were experimental, but never really usable.

Note that there will not be that many features in the 2.0 branch, but the focus is rather on better code architecture, better UX, extraction of modules and removal of features that don't work well.

By removing modules, we hope we can have an even faster development and release pace going forward.

Check the issue queue for more information.

Multiple Automators on one field

One of the larger rewrites for 2.0 is Automators - and one of the features this will bring is the possibility to add multiple Automators one field. While this has been possible in theory in the 1.x branches, this required you to do complex config imports - now they will be a part of the native UI/UX of the Automators.

This together with the Field Widget Actions, means that you can have multiple ways of generating one part of a field, or one button per part of a field.

This means that on an image field for instance, you could setup a button that generates the image, another that renames the file name and a third that creates an alt text.

Or if you want a summarize button in a friendly manner and another in a formal manner, you can have both on the same field.

This is ready to be tested on 2.0.x-dev, but please note that the external Automators modules will also need changes due to breaking changes.

A huge thanks to Bryan Sharpe from ImageX for this and anyone helping with testing and reviewing.

Across every sector, leaders see the potential of AI, but many are struggling to turn that potential into measurable value.

Gartner's 2023 and 2024 findings show that a significant proportion of AI projects stall before reaching production, often due to unclear use cases and limited organizational readiness. The MIT Sloan Management Review highlights that, while most executives recognize AI's transformative potential, few can translate it into ROI.

That's why we created our new series of Drupal AI Industry Guides, to help organizations understand where AI can deliver real, responsible, and measurable impact. Each guide is written by sector experts with firsthand experience implementing AI with Drupal.

Designed for Action, Not Theory

These guides are about doing more with less - a key message in today's digital landscape. Each one features four practical AI use cases that your organization can adopt right now using Drupal AI.

Every use case is grounded in ROI and designed to show how AI can:
Simplify and speed up workflows

• Improve customer and citizen experiences
• Unlock data-driven insight
• Reduce operational costs and risk

Drupal AI brings freedom of choice and rapid innovation to AI adoption. It combines the trust of open source with the power of artificial intelligence.

Expert-Led Guides for Every Industry

Each guide is authored by a Drupal AI sector specialist, professionals who understand both the technology and the real-world challenges of their industries. These guides are designed to move teams from curiosity to capability, showing that

AI doesn't have to be experimental - it can be effective, ethical, and immediate.

Explore the collection:

• Arts and Culture
• Enterprise
• Financial Services and Insurance
• Fintech
• Healthcare and Life Sciences
• High Tech
• Higher Education
• Local Government
• Media and Publishing
• Membership Bodies
• NGO and Nonprofit
• Public Sector
• Retail and Ecommerce
• Sports and Entertainment
• Travel

Responsible, Open, and Future-Ready

Every Drupal AI solution is built on the freedom of open source - giving you control, transparency, and flexibility to experiment safely. Drupal AI is:

• Trustworthy by design: safeguards and auditability built in.
• Human-centered: AI amplifies human creativity rather than replacing it.
• Open by default: integrate today's best AI models and tomorrow's breakthroughs without lock-in.

Drupal is the best AI-powered open-source CMS in the world - built for innovation, designed for trust.
From Curiosity to Capability

The AI opportunity is real, and with Drupal, it's achievable right now. The Drupal AI Industry Guides are your starting point to:

• Identify where AI creates genuine valu
• Build a roadmap for measurable ROI
• Move from pilot to scale with confidence

By combining Drupal's open web foundation with responsible, human-centered AI, you can deliver the innovation and productivity of AI while preserving the judgment, creativity, and context that only people can bring.

AI Makers are a select number of Drupal Certified Partners who are funding and actively working on Drupal AI to ensure it delivers a significant and strategic impact. They provide consulting services to help you build the best AI based digital experience for your needs.

Start Exploring Today

Visit the full collection of Drupal AI Industry Guides and discover how Drupal AI can help your organization move faster, innovate responsibly, and stay open to what's next.