Knowledge Graph
The knowledge graph is the core of ScholarOS — a living map of everything you know. It connects your notes, extracts relationships, and surfaces insights you didn't even know you had.
What Is the Knowledge Graph?
The knowledge graph is a structured database that sits inside your vault. It's powered by SQLite and organises your knowledge as a tree of interconnected nodes.
Nodes
Every concept, topic, or idea in your vault becomes a node in the graph. Nodes have titles, descriptions, and connections to other nodes. Each node maps to one or more files in your vault.
Edges
Edges define relationships between nodes: “is a prerequisite of”, “builds upon”, “contradicts”, “is an example of”. The AI detects these relationships automatically from your content.
Tree structure
The graph is organised as a tree with three fixed root branches. Every node lives under one of these roots, creating a clean, browsable hierarchy of your knowledge.
Local storage
The entire graph is stored in a single SQLite file inside your vault. No data is ever sent to a server. The graph is yours, forever.
How the Graph Builds Connections
Every time you add a document or create a note, the AI processes the content and updates the graph. Here's how it works:
Concept extraction
The AI reads your document and identifies key concepts, terms, definitions, and named entities. Each becomes a candidate node in the graph.
Relationship detection
The AI analyses how concepts relate to each other within the document and across your vault. It creates edges with typed relationships — not just “connected to” but “is a prerequisite of” or “provides evidence for”.
Graph merge
New nodes and edges are merged into the existing graph. If a concept already exists, the new information strengthens the existing node rather than duplicating it.
Wiki page generation
For important nodes, the AI generates or updates a wiki page in your vault. The wiki page becomes a permanent reference that you can read, edit, and link to from other notes.
The Three Fixed Branches
Every node in the knowledge graph lives under one of three root branches. This structure keeps the graph organised and easy to navigate.
User Branch
Your personal knowledge — notes you've written, annotations, summaries, and reflections. This branch represents your understanding and perspective on the material. The AI may suggest additions, but you control what goes here.
Directives Branch
Instructions and guidance — study plans, syllabus requirements, exam specifications, and your learning goals. The AI uses this branch to tailor its outputs to your specific academic context.
World Branch
External knowledge — facts, concepts, and ideas extracted from your textbooks, lecture slides, and past papers. This is the canonical knowledge that the AI synthesises from your source material.
Nodes can link across branches. For example, your personal understanding of a concept (User) might reference a textbook definition (World) and connect to an exam topic (Directives). The graph captures all of these relationships.
Viewing the Graph
ScholarOS provides an interactive graph view that visualises your knowledge network. You can access it from the sidebar or by pressing Cmd+Shift+G.
Graph visualisation
The graph view renders nodes as circles and edges as connecting lines. Node size reflects importance (how many other nodes reference it). Colour indicates which branch the node belongs to. You can pan, zoom, and drag nodes to explore the structure.
Search and filter
Use the search bar to find specific nodes or filter by branch. You can also highlight the shortest path between two nodes to see how concepts are indirectly connected.
Click to inspect
Click any node to open its details panel. You'll see the node description, connected files, related wiki pages, and a list of all edges. From here you can jump directly to the relevant note or wiki page.
Knowledge Compounds Over Time
The knowledge graph gets smarter with every addition. Unlike traditional note-taking where notes sit in isolation, the graph actively grows and strengthens.
Density increases
Each new document adds nodes and edges, but it also strengthens existing connections. The graph becomes denser and more connected over time. A vault with 100 documents is exponentially more useful than one with 10.
Cross-semester connections
As you add material from different courses and semesters, the AI discovers unexpected connections. A concept from a second- year course might link back to a first-year foundation topic, reinforcing your understanding of both.
Spaced repetition
The graph tracks how often you access each concept and when it was last reviewed. It can surface older topics for revision at optimal intervals, helping you retain knowledge long after the exam.
Privacy & Data
All data is local
The knowledge graph is stored as a SQLite database file inside your vault folder. No part of the graph — not the nodes, not the edges, not the relationships — ever leaves your machine. There is no cloud sync, no telemetry, no data collection.
No network access
The graph builds itself purely from your local files. The AI models process your content to extract concepts and relationships, but the resulting graph structure stays on your disk. Even the web search toggle is off by default and clearly labelled when active.
Portable and durable
Because the graph is stored in an open SQLite format, you can query it directly with any SQLite tool, back it up with your regular backup system, or take it with you wherever you go. If you ever stop using ScholarOS, the graph file remains readable on any system.
Start Building Your Graph
Ready to see your knowledge take shape? Start adding your notes and watch the graph grow.