agenticorglab/teachmaster — explained in plain English
Analysis updated 2026-05-18
Study how a multi-agent system can turn a lesson outline into a finished teaching video.
Explore using executable code as an editable middle step in video generation research.
Reference the ACL 2026 paper results when researching AI-assisted educational content.
Evaluate the approach as a lower-cost alternative to traditional course video production.
| agenticorglab/teachmaster | 0xdea/ttyinject-rs | 0xhossam/uncanny | |
|---|---|---|---|
| Stars | 12 | 12 | 12 |
| Language | — | Rust | C |
| Setup difficulty | hard | moderate | hard |
| Complexity | 4/5 | 4/5 | 5/5 |
| Audience | researcher | researcher | researcher |
Figures from each repo's GitHub metadata at analysis time.
The README is paper and results focused with no clear installation or usage instructions provided.
TeachMaster is a research project, presented at the ACL 2026 conference, that generates full educational videos from a teacher's lesson outline and goals. The idea behind it is called Generative Teaching: the educator describes what should be taught at a high level, and a set of AI agents handle the detailed work of planning the lesson, designing visuals, rendering animation, writing narration, converting narration to speech, and checking the result for quality. What sets this apart from typical AI video generation is that TeachMaster does not generate raw video pixels directly. Instead it produces executable code as a middle step, which describes the layout, timing, and animation of each scene. Because the output is code rather than a finished video file, it stays easier to inspect, edit, and adjust before the final video is rendered, and a teacher can either describe changes in plain language or edit the underlying code directly. The system breaks video creation into three stages: content planning, which turns a teaching intent into a structured written blueprint, presentation generation, which turns that blueprint into animated visuals, narration text, and synthesized speech, and quality validation, which checks that the audio and visuals stay synchronized and that the layout renders cleanly, fixing problems automatically where possible. According to the README, TeachMaster has been used by more than 1,500 educators to produce over 140,000 minutes of teaching content across more than 420 subject areas, and the authors report that producing a full semester length course this way costs a small fraction of traditional video production. This project is aimed at researchers and developers working on AI-assisted education, rather than at people looking for a ready-to-use consumer app, since the README focuses on the research paper and results rather than installation steps.
A research system that turns a teacher's lesson outline into a full educational video using AI agents and code as an intermediate step.
Setup difficulty is rated hard, with roughly 1day+ to a first successful run.
Mainly researcher.
This repo across BitVibe Labs
double-check against the repo, no cap.