Analysis updated 2026-05-18
Turn a paid video course into an AI tutor that knows the material
Ask your agent which lesson covers a specific concept and get the timestamp
Apply an instructor's decision framework to your own situation
Build a searchable knowledge base from course videos, PDFs, and notes
| juneyaooo/lineage-skill | hamid-k/nginx-rift-private-lab | nvlabs/spatialclaw | |
|---|---|---|---|
| Stars | 67 | 67 | 67 |
| Language | Python | Python | Python |
| Setup difficulty | moderate | hard | hard |
| Complexity | 3/5 | 5/5 | 5/5 |
| Audience | vibe coder | researcher | researcher |
Figures from each repo's GitHub metadata at analysis time.
Requires API keys for speech-to-text, vision, and text models depending on your source materials.
Lineage Skill is a tool for turning a video course or training program into a custom AI agent skill. The idea is that buying or completing a course is not the same as retaining the instructor's judgment frameworks, case examples, and decision criteria. This project provides a pipeline that processes the raw course materials (video files, audio, PDF handouts, screenshots, transcripts, and personal notes) and extracts and organizes that knowledge into a format an AI agent can use to behave like a tutor who actually knows the course. The pipeline follows six steps: capturing source materials, preserving citations back to the original lesson or timestamp, compressing long materials into structured summaries, connecting related concepts and cases, encoding the instructor's methods into checklists and templates, and then evaluating responses against the course's own standards. The output is an installable agent skill that lets you ask the agent things like "does this conclusion appear in the course, and if so where?" or "apply the framework from this course to my specific situation." The tool is designed for AI coding agents such as Codex or Claude Code. You install the skill by sending a short prompt to your agent pointing it at the install document. You then point the agent at your course materials folder. Depending on what you have, you may need API keys for a speech-to-text service (for audio and video), a vision-capable model (for understanding slides and diagrams), and a text model for the distillation and structuring work. Existing transcripts or notes can skip the audio and video processing steps. The README is written primarily in Chinese. The project is licensed under PolyForm Noncommercial 1.0.0, meaning personal and research use is allowed but commercial use requires contacting the author. One example skill built using this pipeline (based on a traditional Chinese medicine course with over 100 GB of video material) is linked from the README.
A pipeline that turns a video course's materials into an installable AI agent skill so an agent can answer questions like a tutor who knows the course.
Mainly Python. The stack also includes Python.
PolyForm Noncommercial 1.0.0, free for personal and research use, but commercial use requires contacting the author.
Setup difficulty is rated moderate, with roughly 1h+ to a first successful run.
Mainly vibe coder.
This repo across BitVibe Labs
Verify against the repo before relying on details.