Most note-taking apps give you a better filing cabinet. A knowledge graph gives you a second brain you can actually interrogate — and metapad lets you build one, tailor it to how you think, and talk to it. Here is why, how, and a starter template you can fork today.

At a Glance

Most of us keep notes. Few of us can use them. They pile up across a notes app, a documents folder, a handful of half-adopted tools — each one internally tidy, all of them mutually disconnected. The promise of a "second brain" is a place where your ideas don't just sit, but connect — where thinking about one thing leads you to the three related things you'd forgotten. That promise is hard to keep with folders and search alone, because the value of a second brain is in the connections, and a folder can't hold a connection.

A knowledge graph can. In this post we make the case for building your second brain as a graph — notes, topics, projects, journal entries and tasks, all linked — and show how Metapad lets you do it: a flexible model you shape to your own thinking, Markdown notes, tags, attached files, powerful search, a read-only mode for sharing, and an AI assistant that answers questions by walking your graph rather than guessing from prose. We finish with a starter template you can fork and make your own in an afternoon.

Your Notes Are Everywhere and Nowhere

You have read the good book. You capture ideas. You have a notes app on your phone, a folder of PDFs you meant to read, highlights in a reading app, a task list somewhere else, and a journal you keep in fits and starts. Each is fine on its own. Together they are a problem.

Because the thing you actually want from all that captured material is almost never "find the note I wrote." It is:

"What was I thinking about spaced repetition last winter — and did any of it turn into something I actually did?"

That question spans a journal entry, two notes, and a task. No folder answers it. Search gets you closer — it finds the words "spaced repetition" — but it can't tell you that the idea in your journal became a note, which inspired a habit you're now tracking. That thread was never written down in any single place. It only exists once the pieces are connected.

This is the quiet failure of the filing-cabinet model of note-taking. We organise for storage — which folder does this go in? — when what we need is retrieval by association, the way memory actually works. Your notes are everywhere, and the connection you want is nowhere.

Why a Knowledge Graph Makes a Good Second Brain

The people who thought hardest about personal knowledge worked this out long before software. Niklas Luhmann's Zettelkasten — a slip-box of tens of thousands of paper notes that he credited with a prolific career — had one core rule: don't file notes under headings, link them to each other. The German sociologist wasn't building an archive; he was building a network he could have a conversation with.

Two principles do most of the work, and both are about structure, not tooling:

Atomic notes. Each note holds one idea. A note about one idea is small enough to reuse, and — crucially — small enough to link precisely to another. A ten-page document can only be filed; a single clear idea can be wired to every other idea it touches.
Links over hierarchy. Knowledge isn't a tree, it's a network. The moment you stop asking "which folder?" and start asking "what does this connect to?", your notes stop being a pile and start being a map.

A knowledge graph is exactly this idea made computable. It represents information as nodes (things — a note, a topic, a project, a journal entry, a task) and edges (typed relationships between them — this note links to that one; this task belongs to that project; this idea was mentioned in Tuesday's journal). Unlike a folder, you can follow the relationships. Unlike plain search, you can ask questions about the structure itself.

A folder tells you where something is. A graph tells you what it's connected to. Only one of those is a second brain.

And once the structure is explicit and typed, something else becomes possible: you can put questions to it that no document could answer — which of my notes are most connected? which tasks trace back to an idea I journalled about? what have I got in "In Progress" that no longer connects to any live project? The graph is not just storage you search; it is a model you can query. That is the difference between a second brain you feed and a second brain you can think with.

This isn't an argument against your existing notes app, by the way — a fast place to capture a thought is worth having. The graph is what you build on top of capture, when you want the connections to start paying you back.

What Metapad Brings to It

Metapad is a platform for building models of complex systems as knowledge graphs — usually organisations, but the same machinery works beautifully at personal scale. Here is what matters for a second brain, walked against a real template (Section 4) you can open and fork.

A flexible semantic model you shape to your own thinking

This is the one that matters most, so it comes first. In most note apps, the shape of your knowledge is fixed by the vendor: you get notes, maybe tags, take it or leave it. In metapad, you design the shape. The starter template models five kinds of thing — Note, Topic, Project, Journal Entry, Task — and the relationships between them. But that's a starting point, not a cage.

Want to manage your own growth? Add an Objective type and its Key Results, then wire the notes, journal entries and tasks that actually move each one — an OKR system that lives inside your second brain rather than in a separate spreadsheet you forget about. Tracking people? Add a Contact. Running a research habit? Add Source and Claim and wire evidence to arguments. Your second brain should fit the contours of your mind and your work — and, just as importantly, it should be able to grow and change as those needs do, without you starting over. The model is yours to evolve. That is what "methodology-agnostic" means in practice: metapad doesn't tell you how to think.

Markdown-based notes

Notes are written in Markdown — the same lightweight, portable formatting you already use for headings, lists, links and emphasis. Your content stays readable, structured, and never locked into a proprietary blob.

Tagging

Alongside the typed relationships, every note can carry tags — zettelkasten, learning, systems-thinking. Tags give you a lightweight, cross-cutting second axis: relationships model how ideas connect, tags model what they're about. Use both.

Assets — PDFs, images and more

A second brain isn't only text. metapad lets you attach assets — the PDF of the paper behind a note, a diagram, a screenshot, an image — so the source material lives next to the idea it inspired, not in a separate downloads folder you'll never open again.

Powerful search and filtering

Sometimes you do just want to find something. metapad gives you search and filtering across the whole graph — by type, by tag, by property, by text. Show me every Task that's In Progress. Every Note tagged memory. Every Journal Entry from January. Filters turn a growing graph into something you can still navigate at a glance.

Browser-based, including on touch devices

metapad runs in the browser — nothing to install, your second brain is wherever you are. It works on touch devices, so you can capture a thought on a tablet on the sofa and pick it up on a laptop the next morning against the same living model.

Read-only mode — share a view, safely

You can publish a read-only view of a model. Share your reading notes with a colleague, hand a curated topic to a study group, or embed a public "digital garden" — and let people explore the graph by following links, without any risk to the original. (It's also exactly how you'll explore the template in the next section.)

An AI assistant that walks your graph

Every metapad model comes with an AI assistant — and, because each model exposes a structured MCP endpoint, you can also talk to your second brain from the AI tools you already use, such as Claude or Claude Code. But here's the part that matters: the assistant answers by reasoning over your typed graph, not by guessing from a wall of text.

Ask "what was I thinking about spaced repetition, and what did it lead to?" and it can traverse the actual relationships — the journal entry where the idea surfaced, the note it became, the habit it inspired — and give you an answer you can trace back to specific nodes. It can also report: summarise a topic, list every open task on a project, surface notes that are richly connected or orphaned and forgotten.

This is worth contrasting with the now-familiar approach of pointing a chatbot at a folder of documents (a technique called retrieval-augmented generation, or RAG). That's genuinely useful for pulling the right paragraph out of a long document — and in practice you want both. But RAG reads text; it can't tell you that Tuesday's journal idea became a note that inspired a task, because that connection was never written in any one document. The graph knows, because the connection is part of its structure.

Search finds your words. RAG finds your paragraphs. A computational knowledge graph answers questions about how your thinking connects — and lets an AI answer them with you.

A Starter You Can Fork

Reading about a graph only gets you so far. So we built a small, opinionated Second Brain template you can open, explore, and fork as the foundation for your own.

Explore the live template (read-only): Second Brain on metapad Follow the links between notes, open a journal entry, trace a task back to the idea that inspired it.

The template models five kinds of thing and the relationships between them:

Node type	What it holds	Example in the template
Note	One atomic idea, in Markdown, with tags	"Zettelkasten Principle: Atomic Notes"
Topic	A higher-level area that gathers related notes	Personal Knowledge Management, Cognitive Science
Project	A goal that groups tasks and notes	Build Second Brain System
Journal Entry	A dated reflection, with mood and weather	"2026-01-15 — read about the Zettelkasten method today…"
Task	Something to do, Kanban-style	Research Zettelkasten tools, Daily review habit

The point isn't the nodes — it's how they're wired together, because that's what makes it a brain rather than a bag:

Notes link to other notes with a reason attached: Atomic Notes → Links Over Hierarchy → Emergence in Complex Systems, the Zettelkasten idea demonstrating itself.
Notes belong to topics; the Cognitive Science and Personal Knowledge Management topics are themselves related to each other.
Tasks belong to a project and depend on each other in a chain — Research tools → Create templates → Connect old notes → Write blog post.
Notes and tasks are mentioned in journal entries, so a passing thought on a Tuesday stays tethered to what it became — the "overwhelmed by the backlog" entry links straight to the Review and connect old notes task it prompted.
A habit you track was inspired by a note: Daily review habit ← Spaced Repetition for Learning.

Open it, ask the AI assistant "which notes are most connected?" or "what has the journal inspired?", and watch it walk those edges. Then fork it and make it yours: keep what fits, delete what doesn't, and — this is the fun part — change the model itself. Add the node types your life actually has.

5. The Same Pattern, at Two Scales

If this feels familiar, that's the point. A second brain built as a computational knowledge graph is the personal-scale version of something we build with organisations every day.

At personal scale, it's a PersonalOS: journal, notes, tasks and the links between them — one connected model of how you think and work. At organisational scale, the same pattern becomes a workOS: strategy, operations, systems and measurement in one connected model of how a company runs. (We wrote about that in Literate Models: How Metapad Runs Your workOS.) Different domain, same rules: structure and narrative in one source, projections rendered from it, and AI grounded in a typed graph rather than a pile of prose.

Learn the pattern on your own notes, and you've learned the pattern that scales to an enterprise. That's not a coincidence — it's the whole idea.

6. What's Next

The problem we started with — notes that are everywhere and nowhere — isn't solved by a better filing cabinet or one more app. It's solved by building your second brain as a graph: atomic ideas, meaningful links, and a model you can question.

Want to try it? Fork the Second Brain template and start with a structure that works, then bend it to your own thinking.

Want to build from scratch? Metapad is free to start — design your own model, write your first notes, and connect them. Your second brain runs in the browser, on any device.

Curious how this scales? The same platform builds enterprise digital twins — see what Metapad does for organisations.

A second brain isn't where you keep your ideas. It's where your ideas keep finding each other.

A Second Brain on Metapad: A Zettelkasten TemplateYou Can Fork