← Latest papers
💻 computer science

From Tools to Teacher-Built Teammates: No-Code Pedagogical Plugin Authoring with LearnAdapt Agentic Studio and PedOS 1.1 Lumina

This paper introduces LearnAdapt Agentic Studio on PedOS 1.1 Lumina, a no-code platform that enables teachers and researchers to create, preview, and deploy customized educational AI plugins through plain English prompts while ensuring safety and controlled telemetry.

Original authors: Nizam Kadir

Published 2026-07-14
📖 4 min read☕ Coffee break read

Original authors: Nizam Kadir

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a world where the AI tools in your classroom aren't just rigid, pre-packaged robots that teachers have to use exactly as they come out of the box. Instead, picture them as a team of digital assistants that teachers can actually build and customize themselves, without needing to know a single line of computer code. That's the big idea behind a new system called LearnAdapt Agentic Studio and its operating system, PedOS 1.1 Lumina.

Think of the current situation like a school where the only allowed games are the ones the principal bought from a catalog. Teachers can't change the rules or add new levels, even if they know their students would learn better with a different challenge. This paper argues that teachers shouldn't just be players stuck with fixed tools; they should be the architects of their own "AI teammates."

Here's how the magic happens, step-by-step:

1. The "Magic Prompt" Kitchen
In this new system, a teacher doesn't need to be a programmer. They just sit down at the Agentic Studio and type a simple request in plain English, like, "I need a science quiz that asks students to guess the answer before showing them the right one." It's like handing a recipe to a super-fast, invisible team of chefs.

2. The Invisible Chef Team
Once the teacher hits send, a whole crew of AI agents (digital helpers) springs into action. One agent checks the recipe for safety, another designs the look and feel, and a third builds the actual game. They work in parallel, like a high-speed kitchen where everyone has a specific job. They turn that simple sentence into a working "plugin"—a small, self-contained piece of software that can be added to the classroom system.

3. The Safety Inspector
Before this new game can be played by anyone, it has to pass a strict safety check. The system runs automated tests to make sure the code is clean and won't break the school's computers. Then, a human administrator (like a head chef) reviews the work. They look at the preview, check the safety scores, and give the final thumbs-up. Only after this approval does the plugin get published to the Plugin Directory, where other teachers can find and install it.

4. The "Secret Diary" Rule
One of the coolest and most important parts is how the system handles data. The paper explains that the system is very careful about what it records. If a teacher is just previewing the plugin, the system uses fake data (mock telemetry) so nothing real is saved. Real data is only collected when a verified teacher is actually using an approved plugin with real students. It's like a diary that only opens its lock when the right key is turned, ensuring privacy and control.

What This System Is (and Isn't)
The authors are clear about what they have achieved. They have built a working system that successfully takes a teacher's idea, turns it into a safe, working plugin, and manages the data collection. They have shown the whole process works, from the first prompt to the final installation.

However, they are also very honest about what this paper doesn't do. They explicitly state that this submission does not claim that using these teacher-built tools actually improves student learning scores yet. They haven't proven that students learn better; they have only proven that teachers can build these tools safely and easily. The next step, which they are planning for a workshop in June 2026, is to sit down with teachers to figure out exactly what kinds of plugins they want most and what kind of evidence they need to see.

In short, this isn't a finished revolution in student grades, but it is a working prototype for a new kind of school: one where teachers aren't just users of technology, but the creators of their own educational teammates.

Drowning in papers in your field?

Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.

Try Digest →