Using ScholarOS

Vault Structure

Your vault is a regular folder with a clear three-layer structure. Understanding this layout helps you organize material effectively and get the best results from the AI.

my-vault/
  raw/
    ├── CS301/
    │   ├── lecture-01-introduction.md
    │   ├── lecture-02-sorting.md
    │   ├── lecture-03-trees.md
    │   └── textbook-chapter-4.pdf
    ├── past-papers/
    │   ├── 2024-exam.pdf
    │   ├── 2025-exam.pdf
    │   └── 2026-exam.pdf
    └── my-notes/
        └── week-3-summary.md
  wiki/
    ├── sorting-algorithms.md
    ├── binary-trees.md
    ├── time-complexity.md
    ├── hash-tables.md
    └── graph-traversal.md
  outputs/
    ├── revision-cs301.md
    ├── mock-exam-2026.md
    └── flashcards-sorting.md
  config.yaml

Raw Sources

Your original documents. This layer is append-only — you add files here and the AI reads them. Nothing in this folder is ever modified by the AI.

What to put here

Lecture notes (markdown)
Textbook excerpts and chapters
Past exam papers (PDF)
Research papers and articles
Your own summaries and annotations

How it works

Organize by subject or topic (folders are flexible)
The AI scans this folder for new or changed files
Changes here trigger wiki updates
You can subfolder however you like

Wiki Pages

AI-generated topic pages. Each concept from your raw sources gets a structured markdown file. These pages are continuously edited by the AI as new information arrives.

What a wiki page looks like

---
title: Sorting Algorithms
sources
  - raw/CS301/lecture-02-sorting.md
  - raw/CS301/textbook-chapter-4.pdf
related
  - time-complexity.md
  - hash-tables.md
---

# Sorting Algorithms

Sorting algorithms arrange elements in a specific order. The choice of algorithm affects time complexity, space usage, and stability.

## Key Concepts

### Comparison-based sorting
Algorithms that sort by comparing pairs of elements. Lower bound is O(n log n) for comparison-based sorts.

## From your notes

> "Quick sort is the most commonly tested sorting algorithm in CS301 exams. It appears in 73% of past papers."

Wiki page features

Front matter with title, sources, and related topics
Bidirectional links to other wiki pages
Citations back to specific raw sources
Continuously updated as new material arrives

Outputs

Generated study materials. These are the artifacts you actually use for revision and practice. Generated on demand from your wiki pages.

Revision Docs

Condensed study guides covering key topics from your vault. Focus on what matters most for your exams.

Mock Papers

Practice exams designed in the style of your actual past papers. The AI learns question patterns from your vault.

Flashcards

Auto-generated flashcards from your notes and wiki. Test yourself on the exact material you need to learn.

Configuration

A config.yaml file in your vault root controls wiki behavior. Most users never need to touch it — the defaults work well.

# config.yaml
name: My CS301 Vault
model: gpt-5
auto_ingest: true
wiki_depth: 2
max_pages: 200

Tips

Start small — add a few lectures and see how the wiki compiles before dumping everything in.

Name files descriptively — the AI uses filenames as signals for topic extraction.

Past papers are gold — the more you add, the better the mock papers become.

Organize by subject — use folders to separate courses (e.g., raw/CS301/, raw/MATH201/).

Don't edit wiki pages manually — the AI manages them. If you need to add notes, add them to raw/.

Need help?

Check the FAQ or reach out on GitHub.

FAQ Features