Documentation

Learn Kontext

Everything you need to know to build structured context for AI.

Quick Start

New to Kontext? Get up and running in 5 minutes.

Open App

Getting Started

Learn the basics of Kontext and create your first canvas.

Core Concepts

Deep dive into the fundamental concepts that power Kontext.

AI Features

Harness the power of AI to expand and refine your ideas.

Export & Share

Share your structured context with AI agents and collaborators.

What is Kontext?

Kontext is a node-based context builder designed for the AI era. Instead of writing traditional documents meant for human readers, you build structured context that AI can natively understand and process.

Think of it as the bridge between your thoughts and AI agents. You pour your ideas onto an infinite canvas, organize them visually, and use AI presets to expand, challenge, or refine your thinking. The result is a Context Packet—a structured bundle of information that any AI can immediately work with.

Node Types

Text Node

Your raw thoughts, notes, or any text content. The building blocks of your canvas.

AI Assistant

Write a prompt and connect input nodes. The AI generates output based on your context.

AI Presets

Instead of writing prompts from scratch, use presets to quickly transform your nodes:

Expand

Add missing perspectives using MECE principles

Deepen

Analyze underlying assumptions and implications

Switch

View from a completely different angle

Challenge

Get constructive counterarguments

MCP Integration

What is MCP?

MCP (Model Context Protocol) is a communication standard between AI assistants and external tools, created by Anthropic. By installing the Kontext MCP server, you can query Context Packets directly from Claude Desktop, Cursor, VS Code, and more.

AI automatically understands your project context without copy-paste.

Available Tools

get_context

Retrieve Context Packets. Get the latest packet or specify a particular ID.

list_packets

List available Context Packets.

list_canvases

List your Kontext canvases.

get_constraints

Get project constraints.

get_decisions

Get decisions derived from canvases.

get_insights

Get AI-analyzed insights.

Installation

Visit the Connect page to set up the MCP server.

Usage Example

Try saying this in Claude Desktop:

"Use the Kontext tool to get my latest context"

Need Help?

Join our community or reach out for support.

CommunityContact Support