zettelkasten

Store and retrieve atomic notes with semantic connections in SQLite.

2|Updated Feb 6, 2026

Search Tags:#sqlite #notes #linking #knowledge-management #semantic-search #zettelkasten

Ready to run

npx skills add https://github.com/eddmann/jeeves --skill zettelkasten-eddmann

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: zettelkasten
Source: https://github.com/eddmann/jeeves/tree/main/skills/zettelkasten
Command: npx skills add https://github.com/eddmann/jeeves --skill zettelkasten-eddmann

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires sqlite-utils>=3.36, sqlite-vec>=0.1, openai>=1.0, and includes scripts (resource) components.

What problem does it solve?

The Zettelkasten Skill helps you store atomic notes, connect ideas with links, and build a searchable personal knowledge base, making it easier to capture and retrieve knowledge over time.

Core Features & Use Cases

Atomic notes: each note captures a single idea to keep concepts modular and composable.
Linking and graph: create links between notes to reveal relationships and build a navigable knowledge graph.
Search and memory: full-text search and optional embeddings enable fast recall and semantic discovery.

Quick Start

Initialize the Zettelkasten database and begin creating, tagging, and linking atomic notes.

Frequently Asked Questions about zettelkasten

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I build a personal knowledge base with atomic notes and semantic search?▼

Build a personal knowledge base by storing atomic notes in a lightweight SQLite store, using full-text search and tag indexing for fast retrieval and semantic discovery. You create, update, delete, and link notes to form a navigable knowledge graph.

How does semantic search work for retrieving linked notes?▼

Semantic search for retrieving linked notes works by generating optional vector embeddings via sqlite-vec, allowing you to query the SQLite store for conceptually similar ideas rather than relying solely on exact keyword matches.

Do I need SQLite and OpenAI to use semantic search for my notes?▼

Yes, semantic search for your notes requires a SQLite database using sqlite-utils and sqlite-vec, along with the OpenAI library to generate the optional vector embeddings needed for semantic queries.

What is the best way to connect ideas and build a knowledge graph in a local database?▼

The best way to connect ideas and build a knowledge graph locally is capturing single atomic notes, then creating links between them in a SQLite database to reveal relationships and enable navigable memory recall across sessions.

Can I use this knowledge management approach without generating vector embeddings?▼

Yes, you can use this knowledge management approach without vector embeddings by relying on the built-in SQLite full-text search and tag indexing to retrieve atomic notes, though you will lose semantic discovery capabilities.