Lore

Browse your knowledge graph in the browser.
Lore is the read-only frontend for a gbrain knowledge brain — force-directed graph, dashboard, and full-text search. Bring your own gbrain backend.

Quickstart  ·  Deploy  ·  Configure  ·  Contribute

What is Lore?

Lore is a read-only web UI for exploring a gbrain knowledge brain — your team's people, products, docs, and the decisions that connect them. It reads gbrain over MCP and renders a force-directed graph, a dashboard, and hybrid full-text search, so you can see and walk your knowledge instead of grepping it. It never writes.

Features

• Force-directed graph — d3 node-link view with smooth zoom/pan, click-to-filter by type, and connection-walking from any node.
• Dashboard — pages, links, sources, daily activity, top hubs, and recent memories at a glance.
• Hybrid search — title + content search over your gbrain, as you type.
• Pluggable viz modules — drop in a new src/lib/viz/<name>.ts to add a visualization.
• Fail-closed auth — none (dev), HTTP Basic, or Cloudflare Access (JWT-verified). Read-only by design.
• Deploy anywhere — standalone Docker image; one-click to Vercel or Railway.

Quickstart

Prerequisite: a running gbrain backend — point GBRAIN_MCP_URL at its MCP endpoint.

git clone https://github.com/corespeed-io/lore.git && cd lore
cp .env.example .env        # set GBRAIN_MCP_URL + GBRAIN_TOKEN — auth is off for local dev by default
npm install && npm run dev  # → http://localhost:3000

Deploy your own

Lore fails closed. A fresh deploy returns 403 until you set AUTH_MODE — proxy (Cloudflare Access) or password, or none with ALLOW_INSECURE=1. It will not serve a private brain by accident.

Lore is a standard Next.js standalone app, so it also runs on Railway (Dockerfile auto-detected) or any container host:

docker build -t lore . && docker run -p 3000:8080 --env-file .env lore

Configuration

Config is entirely environment-driven — see .env.example for the full list.

Variable	Required	Notes
GBRAIN_MCP_URL	yes	Your gbrain MCP server endpoint
GBRAIN_TOKEN	*	Static bearer (server-only, never sent to the browser)
GBRAIN_CLIENT_ID / GBRAIN_CLIENT_SECRET	*	Preferred: a read-only OAuth client — Lore mints short-lived tokens
APP_TITLE / APP_SUBTITLE	no	Hero branding, per deployment
AUTH_MODE	no	none · password · proxy (Cloudflare Access). Defaults to none
ALLOW_INSECURE	no	Required to actually run with AUTH_MODE=none (auth fails closed otherwise)
ACCESS_TEAM_DOMAIN / ACCESS_AUD	for proxy	Cloudflare Access team domain + audience

* Provide either GBRAIN_TOKEN or a client id/secret. A read-only client is recommended so a leaked credential can't write.

Auth fails closed. none is honored only when ALLOW_INSECURE=1 is also set; proxy verifies the Cloudflare Access JWT (signature, audience, issuer, expiry) and denies if it's misconfigured. Never expose the origin with AUTH_MODE=none to the internet.

Extending — add a visualization module

// src/lib/viz/<name>.ts
export function mount<Name>(element: HTMLElement, data: GraphData, options: VizOptions): void {
  // render with d3, canvas, or the DOM
}

Mount it from src/components/GraphView.tsx and add a test in tests/.

Development

npm run dev        # dev server (hot reload)
npm run typecheck  # tsc --noEmit
npm run lint       # biome
npm test           # vitest
npm run build      # production build

Working with an AI coding agent? AGENTS.md is the single source of truth — Claude Code, Codex, Cursor, Gemini, and Copilot all read it.

Contributing

Issues, ideas, and PRs are welcome — start with CONTRIBUTING.md or open a discussion. Built and maintained by CoreSpeed.

License

MIT © CoreSpeed