Launching today
PAIDEIA
Turn course PDFs into a local exam-prep brain
1 follower
Turn course PDFs into a local exam-prep brain
1 follower
PAIDEIA is a local-first Claude Code plugin for students who study from messy course PDFs, homework, solutions, and handwritten mistakes. Instead of teaching an average syllabus, it builds your course-specific exam graph: PDF-to-Markdown ingestion, solution pattern extraction, homework-density weighting, OCR grading, weakmaps, drills, mock exams, and cheat sheets. Every artifact stays on disk as editable Markdown.
Hey Product Hunt!
I’m Taewoo, a physics and mathematics undergraduate at KAIST. I built PAIDEIA after seeing students, including myself, rely on the same pattern before exams: upload a pile of lecture PDFs into a chatbot, ask for summaries, then lose most of the context when the session ends.
Quick summaries are useful, but they miss the signals that actually matter in a real course:
- what your professor repeatedly emphasizes
- which homework patterns keep appearing
- where you personally make mistakes
- which topics are likely to show up on the exam
PAIDEIA is my attempt to make that study state persistent.
It is a Claude Code plugin that turns your course PDFs, homework, solutions, and handwritten scanned answers into a local exam-prep knowledge base.
The workflow is:
ingest → analyze → drill → grade → weakmap → cheatsheet
It can convert PDFs into Markdown, extract recurring solution patterns, rank topics by homework density, generate drills and mock exams, OCR-grade handwritten answers, and build weakmaps and cheat sheets from your own mistakes.
The important part: everything stays on disk as editable Markdown. You can open it in Obsidian, edit it manually, commit it, or keep using it even if the model changes later.
I added real Claude Code terminal demo videos to the README so the workflow is easier to judge.
I’d love feedback from students, educators, Claude Code users, and anyone interested in local-first AI tools.