Use when building a long-term knowledge base from reading, research, or learning that needs to generate new insights rather than just store information
Scanned 9/8/2026
Install to Claude Code
npx -y skills add jeffreytse/grimoire-core --skill apply-zettelkasten --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Apply Zettelkasten?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/jeffreytse-apply-zettelkasten)More formats (shields.io, HTML) on the badges page.
---
name: apply-zettelkasten
description: Use when building a long-term knowledge base from reading, research, or learning that needs to generate new insights rather than just store information
source: Niklas Luhmann, "Kommunikation mit Zettelkästen" (1981); Sönke Ahrens, "How to Take Smart Notes" (2017)
tags: [knowledge-management, note-taking, zettelkasten, research, writing, learning]
verified: true
---
# Apply Zettelkasten
Build a network of atomic, linked notes — one idea per note, connected by explicit links — that surfaces unexpected insights across domains over time.
## Why This Is Best Practice
**Adopted by:** Academic researchers, non-fiction writers, software architects, knowledge workers in law, philosophy, and science using tools like Obsidian, Roam, and Logseq
**Impact:** Luhmann: used his Zettelkasten to produce 70 books and 400 academic articles over a 40-year career; Ahrens: linear note-taking produces archives you never re-read; Zettelkasten produces a thinking partner that generates connections you didn't plan
**Why best:** The core insight is that the value of notes is not in storage but in retrieval and collision. When each note is atomic (one idea, standalone) and explicitly linked to related notes, the act of placing a new note forces you to articulate how it relates to existing knowledge — and that articulation is where insight is generated.
## Steps
1. **Capture fleeting notes while reading** — Quick, rough notes in your own words; shorthand is fine; these are temporary and must be processed the same day or they lose context
2. **Write literature notes** — For each source: record bibliographic info, then summarise the key ideas in your own words; include your assessment and disagreements; one note per source
3. **Write permanent notes** — For each idea worth keeping: one note, one idea, written as if to a future reader who has no context; it must be standalone and complete
4. **Link explicitly** — When filing a permanent note, ask: what other notes does this connect to? Add bidirectional links by ID or title; write a sentence explaining the relationship
5. **File by concept, not source or date** — Permanent notes live near the ideas they relate to, not in a "Books read in 2025" folder; structure emerges from links, not from imposed categories
6. **Follow links when writing** — When starting a writing project, enter the Zettelkasten through a relevant note and follow links; the network reveals the structure of your argument
## Rules
- One idea per note — if a note covers two ideas, split it; compound notes cannot be linked precisely
- Always write in your own words — copy-paste creates an archive; re-phrasing creates understanding
- Fleeting notes must be processed within 24 hours or discarded; stale fleeting notes are noise
- Every permanent note must have at least one link to an existing note; unlinked notes are dead weight
- Never organise by source ("notes on Book X") — organise by idea; the source belongs in the bibliographic note only
## Examples
**Fleeting note:** "Newport — deep work produces in hours what reactive work does in days — attention residue lasts 20 min"
**Literature note:** Newport (2016) argues that cognitive performance depends on extended uninterrupted focus because task-switching leaves attention residue that degrades subsequent work for up to 23 minutes. I partially disagree: this may vary by task type.
**Permanent note (ID: 2025-06-03-attention-residue):** "Attention residue: switching tasks leaves a cognitive echo of the previous task that impairs performance on the new one for 10–23 minutes (Mark 2008). This makes interruption cost asymmetric — the interrupter pays zero cost; the interrupted person pays 23 minutes." Links: → [2025-05-12-context-switching-cost], → [2025-04-20-deep-work-scheduling]
## Common Mistakes
- **Building a reference archive instead of a thinking tool** — Highlighting PDFs and storing them is not Zettelkasten; the note must be written in your own words or it produces no understanding
- **Not linking when filing** — A note without links is isolated; the network effect only emerges from explicit connections
- **Using the tool as a to-do list or project manager** — Zettelkasten is for developing ideas, not tracking tasks; mixing the two degrades both
- **Waiting to write permanent notes until you "understand enough"** — Write the note at the frontier of your current understanding; confusion articulated is the beginning of insight
## When NOT to Use
- When the knowledge domain is short-lived or project-specific (a one-off compliance requirement, a temporary workaround) — investing in permanent linked notes for information with no long-term value creates maintenance debt with no compounding return.
- When you need to produce output immediately — the Zettelkasten method is a slow-build system optimised for long-term synthesis; if a deadline is hours away, writing directly from sources is faster than routing through the note network.
- When the person is a beginner in a new field who has not yet built enough conceptual scaffolding to identify what is worth making atomic — premature atomisation of half-understood ideas produces an unlinked pile of fragments rather than a connected knowledge network; linear note-taking is more appropriate until a baseline mental model is established.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!