Zettelkasten

Zettelkasten

AI-powered Zettelkasten knowledge base exposed via MCP for AI assistants.

8
Stars
3
Forks
0
Releases

Overview

Zettelkasten AI Assistant is an AI-powered Zettelkasten knowledge management system designed to capture, refine, and connect atomic ideas. Built with modern Python, it stores notes as Markdown files with YAML frontmatter and maintains a SQLite index for fast full-text search. Notes are atomic (one concept per note) and uniquely identified, with bi-directional links forming a knowledge graph. The system offers AI-assisted workflows to Capture, Explain, Question, Refine, and Connect, including auto-summarization and metadata suggestions to surface meaningful insights and relationships. Storage is future-proof by design, while AI features help reduce cognitive load and accelerate learning. It exposes multiple interfaces—CLI, REST API, Streamlit UI—and an MCP Server that lets external AI assistants and agents access and manipulate notes via a defined toolset (create, search, CEQRC execution, links, and summaries). OpenAI is supported (with a stub option for testing), and complete documentation guides integration and testing.

Details

Owner
joshylchen
Language
Python
License
Other
Updated
2025-12-07

Features

Atomic Notes with Unique IDs and Metadata

One concept per note with a unique ID, title, body, summary, and timestamped metadata for precise linking.

Bi-directional Linking and Knowledge Graph

Typed relationships between notes form a navigable knowledge graph to enable concept connections.

AI-Powered CEQRC Workflow

Capture → Explain → Question → Refine → Connect pipeline to guide understanding and relationships, powered by AI.

Auto-Summarization and Metadata Suggestions

AI-generated concise summaries and suggested titles, tags, and relationships to surface key insights.

YAML Frontmatter + Markdown Storage

Notes stored as Markdown with rich YAML frontmatter, serving as the source of truth and metadata container.

Full-Text and Semantic Search

SQLite FTS5 index supports full-text search with proximity, phrase queries, and tag-based filtering.

Multiple Interfaces

CLI, REST API, Streamlit UI, and MCP Server provide broad access for use and automation.

MCP Integration and Tools

Expose to AI assistants via MCP; toolset includes zk_create_note, zk_search_notes, zk_get_note, zk_run_ceqrc_workflow, zk_suggest_links, zk_create_link, zk_generate_summary.

Audience

ResearchersCapture research notes, link concepts, and refine ideas with AI workflows.
StudentsSummarize materials, generate study questions, and build connected notes for exam preparation.
Knowledge workersOrganize information, meeting notes, and project knowledge with AI-assisted linking.
AI integratorsExpose knowledge base to external AI assistants via MCP for automation.

Tags

zettelkastenknowledge-managementmcpai-assistantnotesatomic-noteslinkingbi-directionalfull-text-searchfts5markdownyaml-frontmatterCEQRCstreamlitCLIREST APImcp-tools