Knowledge Tree Documentation

Knowledge Tree is a knowledge integration system that builds understanding exclusively from raw external data — never from AI model internal knowledge. It constructs and evolves a knowledge graph where every node is grounded in provenance-tracked facts decomposed from real sources.

What makes Knowledge Tree different

• Knowledge from data, not from models. AI models are reasoning engines, not knowledge sources. All knowledge traces back to external raw data.
• Integration, not ignoring. The system never discards coherent information. Contradictory facts produce perspectives with alternate viewpoints, not suppression.
• Multi-model convergence. Multiple AI models analyze the same evidence independently. Consensus reveals genuine truth; divergence reveals where biases determine conclusions.
• Transparent provenance. Every claim traces through facts back to original sources. Nothing is hidden.
• Accumulation. The graph improves with every query. Frequently explored topics become deeply supported over time.

Documentation sections

How It Works

Understand the system's core concepts — from facts and entities to seeds, nodes, dimensions, and synthesis. Learn how Knowledge Tree builds a living knowledge graph from raw sources.

MCP Integration

Connect your AI tools to the Knowledge Tree graph via the Model Context Protocol. Browse nodes, facts, dimensions, and relationships directly from Claude Desktop or any MCP client.

Contributing

Dive into the architecture, core objects, services, and development setup. Everything you need to start contributing to Knowledge Tree.

Services

Service	URL	Description
Landing Page	openktree.com	Project overview and links
Research App	research.openktree.com	Ingest sources, create syntheses, explore the graph
Wiki	wiki.openktree.com	Read-only knowledge graph browser
Docs	docs.openktree.com	Developer documentation (you are here)
MCP Server	mcp.openktree.com	Model Context Protocol endpoint
API	api.openktree.com	REST + SSE API