The open platform for knowledge engineering

Engineer your knowledge.
Share it with anyone.

Backpack is the knowledge engineering layer for AI. Build learning graphs that structure what you know. Encrypt and share them privately, or publish to the world. Open source.

Start free Install locally

Build. Share. Explore.

Build a learning graph

Describe any topic to your AI — clients, research, processes, decisions. Backpack structures it into a visual learning graph you can explore and grow over time.

Share securely

Your graph is encrypted on your device before it leaves. Share a private link with your team, or publish it to the graph commons for anyone to discover.

Explore and discover

Browse learning graphs published by the community. Fork them, extend them, or use them as a starting point for your own research.

Explore a learning graph

A real learning graph about semiconductor manufacturing. Pan, zoom, click any node.

Private by default, open by choice

Your data is yours. Share only what you want, with who you want.

Encrypted on your device

Learning graphs are encrypted before they leave your machine. Our servers store the encrypted data and have no way to read it.

Zero-knowledge sharing

When you share a link, the decryption key is embedded in the URL itself. It never reaches our servers. Only people with your link can view the graph.

Publish to the commons

Want to share openly? Publish a learning graph to the public commons where anyone can browse and build on it. Great for open research and community knowledge.

Open source

The graph engine, viewer, sharing protocol, and encryption are all open source. Run everything locally with zero dependencies on our cloud.

What people build

Research

Literature reviews, topic maps, citation networks. Build a learning graph from papers and sources, then let your AI synthesize across them.

Business

Client relationships, competitive landscape, operational processes. Give your AI the full picture so it can help across every conversation.

Community

Publish domain-specific learning graphs to the commons. Let others fork, extend, and connect your work with theirs.

Personal

Project notes, learning journals, decision records. Carry structured context across every AI session — nothing gets lost.

Works with every MCP tool

Built on the open Model Context Protocol. Works everywhere MCP does.

Claude Code

Claude App

Cursor

Windsurf

OpenAI Codex

Cline

Continue.dev

Zed

Get started

Local

Free forever · fully offline

Everything stays on your device. Unlimited graphs, full MCP server, no account needed.

claude mcp add backpack-local -s user -- npx backpack-ontology@latest

Add to ~/.cursor/mcp.json:

{
  "mcpServers": {
    "backpack": {
      "command": "npx",
      "args": ["backpack-ontology@latest"]
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json:

{
  "mcpServers": {
    "backpack": {
      "command": "npx",
      "args": ["backpack-ontology@latest"]
    }
  }
}

codex mcp add backpack -- npx backpack-ontology@latest

Cline — Click the MCP Servers icon in Cline’s top bar and add the JSON config above.

Continue.dev — Add to ~/.continue/config.yaml:

mcpServers:
  - name: backpack
    command: npx
    args:
      - "backpack-ontology@latest"

Zed — Add to ~/.config/zed/settings.json under "context_servers" (not "mcpServers").

Requires Node.js 18+. Works with any MCP-compatible tool.

Cloud

Encrypted sharing · graph commons · sync

Sync learning graphs from your local install to the cloud. Share encrypted links or publish to the commons.

Create free account

1 free encrypted graph + 4 public graphs. See all plans →

Knowledge engineering for everyone.

Open source. Private by default. Free to start.

Start free Read the docs

Engineer your knowledge.Share it with anyone.