Files
zeekay 61644272ea gear-locker: track company equipment on @hanzo/gui + @hanzo/iam + @hanzo/base
Assign and track company assets — who holds each one and its status.
Forked from the canonical hanzo-starter: identical provider stack, vite.config,
PKCE auth against hanzo.id, and static-SPA deploy contract. Three views built on
@hanzo/gui primitives (longhand props): asset list, asset detail (with
assignment timeline), and a global assignments timeline. schema.sql provisions
org-scoped Base collections assets + assignments.

tsc --noEmit clean; vite build green; renders with zero console errors.
2026-07-21 13:23:49 -07:00

5.1 KiB

Gear Locker

Assign and track company equipment. A real, buildable Hanzo app you fork on hanzo.app and deploy live on Hanzo Cloud.

Track company assets — laptops, monitors, phones — who they're assigned to, and their status. Assign gear to a person, log returns, and read the full hand-off history. Every row lives per-org in Hanzo Base.

  • UI — @hanzo/gui (the Hanzo design system) under Vite + React 19. No second kit, no Tailwind.
  • Auth — @hanzo/iam, OAuth2 PKCE against hanzo.id. No local passwords — IAM owns every credential interaction.
  • Data — @hanzo/base, the IAM-native, org-scoped data plane. assets and assignments are real Base collections.

Views

  • Asset list — every asset with a monospace inventory tag, type, status badge, and current holder. Add new gear inline.
  • Asset detail — the selected asset's status, holder, and a full assignment timeline. Assign it to a person, return it, change status, or delete it.
  • Assignments — one chronological timeline of every hand-off across the locker, with the live one flagged current.

Stack (pinned)

Package Version
react / react-dom ^19.2.4
@hanzo/gui + @hanzogui/config 7.3.0
@hanzo/iam ^0.13.1
@hanzo/base ^0.2.1
vite ^6 (@vitejs/plugin-react)
react-native-web ^0.21.0
typescript 5.9.3

Run it

npm install
npm run dev        # http://localhost:5173
npm run build      # tsc --noEmit && vite build  ->  dist/
npm run preview    # serve the production build (SPA fallback on)

Out of the box it runs against live Hanzo (hanzo.id + api.hanzo.ai) — no config needed to see the sign-in flow. Copy .env.example to .env to point at a different environment.

Environment contract

Only VITE_-prefixed vars reach the browser (this is a static SPA — there is no server). Defaults in parentheses.

Var Purpose
VITE_HANZO_IAM_URL (https://hanzo.id) OIDC issuer.
VITE_HANZO_CLIENT_ID (hanzo-app) IAM application (<org>-<app>). Its redirect-URI list must allow this deploy's /auth/callback — see Ambient IAM.
VITE_HANZO_REDIRECT_URI (${origin}/auth/callback) PKCE redirect.
VITE_HANZO_BASE_URL (https://api.hanzo.ai) Browser-reachable Hanzo Base data plane. Deploy injects the provisioned URL.

How auth works — ambient IAM

login() starts an OAuth2 PKCE S256 redirect to hanzo.id; hanzo.id returns to /auth/callback, where handleCallback() exchanges the code for tokens (stored in localStorage, refresh-aware via offline_access). Every deployed app is a static site at <slug>.hanzo.app; there is no server token — the SPA authenticates the user in the browser and carries the resulting IAM JWT to Base. "Ambient" means the app just reads the signed-in user via that token.

The one deploy requirement: the IAM client (VITE_HANZO_CLIENT_ID) must list this origin's /auth/callback as an allowed redirect URI. Register a https://*.hanzo.app/auth/callback wildcard on the shared client so every forked app works, or register a dedicated gear-locker client.

How data works — Base from schema.sql

schema.sql is the app's databaseSchema (SQL DDL). On publish, Hanzo Cloud translates each CREATE TABLE into a Hanzo Base collection (provisionBaseFromDDL, additive + idempotent). Base manages id/created/updated/owner/org, stamps owner+org from the verified IAM principal, and scopes every row to the caller's org (rule @request.auth.org_id = org) — a teammate in your org sees the row; other orgs cannot. At runtime the views read/write assets and assignments through @hanzo/base/react (useQuery/useMutation) carrying the IAM token. Keep schema.sql in lockstep with src/lib/model.ts.

Deploy — Hanzo Cloud

hanzo.yml declares a static build (npm run build → dist/, served at <slug>.hanzo.app) plus the Base schema to provision and the env to inject. Do not build a container image locally — Hanzo Cloud owns builds and deploys. CI here only proves the template compiles green.

Layout

src/
  main.tsx          entry
  providers.tsx     GuiProvider -> IamProvider -> BaseProvider(client=IAM-token)
  app.tsx           route (/auth/callback) + auth gate
  gui.config.ts     createGui(defaultConfig from @hanzogui/config/v5)
  iam.config.ts     IAM PKCE config
  env.ts            the VITE_ env contract, one place
  lib/base.ts       BaseClient carrying the IAM bearer token
  lib/theme.ts      the dark palette + monospace stack, one place
  lib/model.ts      Asset/Assignment types, status vocabulary, pure helpers
  ui/atoms.tsx      the reusable @hanzo/gui design atoms
  auth/callback.tsx PKCE return leg
  views/            signed-out (landing) · home (shell) · assets · assignments
schema.sql          databaseSchema -> Base collections on publish
hanzo.yml           Hanzo Cloud build/deploy manifest