Skip to content
CKS Documentation

Canonical Knowledge Structure

A universal, representation-independent foundation for verifiable AI knowledge.

CKS is an open ecosystem that gives LLMs a canonical knowledge backbone. Every piece of information must be explicitly structured, validated against formal constraints, and traceable to its origin. This eliminates hallucinations and makes AI-generated knowledge auditable.

How It Works

Section titled “How It Works”
Your LLM (Claude Desktop, etc.)      cks-studio (visual workspace)
        │                                     │
        └───────────────┬─────────────────────┘
                         ▼
                     cks-mcp ─── Model Context Protocol server
                         │
                         ▼
                  cks-runtime ─── Sessions, transactions, version history
                         │
                         ▼
                    cks-core ─── Immutable semantic engine

An LLM and a human can drive the same knowledge graph at the same time — one over MCP through Claude Desktop, the other clicking around the graph canvas in cks-studio — because both go through the same cks-mcp tool surface into the same cks-runtime session.

Key Capabilities

Section titled “Key Capabilities”
  • Eliminate citation hallucinations – mechanically detect references to non-existent sources.
  • Ensure verification integrity – cryptographic signing guarantees that source checks actually happened.
  • Full audit trail – every operation is captured in an immutable version history.
  • Time-travel debugging – list versions, compare them, and safely roll back to any previous state.
  • Branch, merge, and sandbox – isolate an experiment from the main line, reconcile it with a three-way merge, or throw it away.
  • Real semantic search – find knowledge by meaning, not keywords, via HuggingFace embeddings.
  • LLM-friendly API – native MCP server with 63 tools, plus MCP Resources and Prompts, fully compatible with Claude Desktop and other MCP clients.
  • Visual workspace – explore graphs, inspect inference chains, review CRDT forks, and monitor agent pipelines from cks-studio, without leaving the browser.

Projects

Section titled “Projects”
Project Description Status
cks-core Semantic engine – immutable knowledge objects, validation, evolution v1.21.1
cks-runtime Operational environment – sessions, transactions, versioning, events v1.49.2
cks-mcp MCP server – exposes CKS to LLMs via 63 tools v1.58.0
cks-studio Visual workspace – graph canvas, gallery, agent monitoring, and an AI chat panel that can call the same tools v0.6.6

Get Started in 5 Minutes

Section titled “Get Started in 5 Minutes”
Terminal window
pip install cks-core cks-runtime cks-mcp

Then connect to Claude Desktop – see the Quick Start guide.

Prefer a visual, click-around interface instead? cks-studio runs alongside cks-mcp and gives you a graph canvas, session gallery, and an in-browser AI chat panel that can call the same tools an LLM would — see its repository for setup.

For a deeper dive into the MCP server specifically: its tools reference, security model, and architecture each have their own page under cks-mcp.

Why CKS?

Section titled “Why CKS?”

Today, the same knowledge exists in many incompatible forms – documents, databases, JSON, source code, AI prompts. CKS separates knowledge itself from every representation. Representations may change, but canonical knowledge remains the same.