Moheeb Zara b8cd36abc3 ci: GitHub Actions auto-deploy to production droplet on push to main
Workflow SSHes into the production droplet as `deploy` and runs
`/opt/hackerspace-ops/deploy.sh` which:
  - git fetches and resets to origin/main
  - rebuilds the app image with NEXT_PUBLIC_* baked in from .env.local
  - applies any new scripts/0NN_*.sql migrations not in the
    public._migrations_applied tracking table
  - hot-swaps the app container with `docker compose up -d --no-deps app`
  - prunes dangling images

Repo secrets DROPLET_HOST and DEPLOY_SSH_PRIVATE_KEY are configured.

Concurrency group `deploy-prod` with cancel-in-progress=false so two
pushes in quick succession queue rather than racing.

Also gitignores `secrets/` and `*.pem` so production credentials never
get committed.
2026-05-14 15:41:33 -07:00
2026-03-11 08:00:22 +00:00
2026-03-11 04:44:48 +00:00
2026-03-11 04:44:48 +00:00
2026-03-11 04:44:48 +00:00
2026-03-11 04:44:48 +00:00
2026-03-11 08:00:22 +00:00

hackerspace-management

Multi-tenant member, payments, tasks, projects, ops, and comms platform for hackerspaces. Each tenant (a "space") owns its members, tasks, projects, payments, knowledge base, and chat channels. All access is row-level secured at the database.

Stack

  • Next.js 16 (App Router, React 19, TypeScript)
  • Tailwind v4 + shadcn/ui
  • Supabase (Postgres, Auth, Realtime)
  • Vitest + Playwright

Quick start (local)

git clone <repo-url>
cd hackerspace-management
pnpm install
cp .env.example .env.local
# Fill in Supabase URL, anon key, service role key in .env.local

Apply the database schema in your Supabase project:

  1. Supabase dashboard, SQL Editor, New query.
  2. Paste the contents of scripts/schema.sql.
  3. Run.

Then:

pnpm dev

Open http://localhost:3000, sign up, create a space, land on /dashboard.

Health probe: curl http://localhost:3000/api/health.

Documentation

File What it covers
CLAUDE.md Working agreement for AI assistants in this repo
docs/AUDIT.md Latest deep audit (2026-05-13): findings, fixes applied, open issues
docs/HANDOFF.md Session-by-session change log
docs/DEPLOYMENT.md Local, Vercel, DigitalOcean App Platform, DigitalOcean Droplet (with managed Supabase)
docs/DEPLOY_DO_SELFHOSTED.md End-to-end self-hosted on a single DigitalOcean Droplet (Supabase + app)
docs/LOCAL_DEV.md Fresh-clone to running app locally in five minutes, using the Supabase CLI
docs/ARCHITECTURE.md System design, project structure, auth flows
docs/DATABASE_SCHEMA.md Full schema reference
DB_SCHEMA_MAP.md Quick column-by-column lookup
docs/API_REFERENCE.md Server actions
docs/COMPONENT_REFERENCE.md UI components
TESTING.md How to run tests
docs/GOVERNANCE_FEATURES.md Diagnostic-to-feature roadmap for governance modules (proposals, incidents, policies)

Deployment

Four supported targets:

  • Vercel (managed Supabase): zero-config, connect the GitHub repo, set env vars. See docs/DEPLOYMENT.md.
  • DigitalOcean App Platform (managed Supabase): spec ships at .do/app.yaml. See docs/DEPLOYMENT.md.
  • DigitalOcean Droplet, app only (managed Supabase): Dockerfile and docker-compose.yml ship in the repo root. See docs/DEPLOYMENT.md.
  • DigitalOcean Droplet, fully self-hosted (your own Supabase, your own Postgres, your own data): full end-to-end guide in docs/DEPLOY_DO_SELFHOSTED.md.

Required environment variables are in .env.example.

Database migrations

scripts/schema.sql is the canonical, idempotent, full-schema deploy. Run it once on a fresh Supabase project. For existing deployments, only run new incremental files (scripts/NNN_description.sql).

Scripts

pnpm dev          # development server
pnpm build        # production build
pnpm start        # serve production build
pnpm lint         # eslint
pnpm test         # vitest watch mode
pnpm test:ui      # vitest UI
pnpm test:e2e     # playwright

Project status

The system is feature-functional but pre-production. See docs/AUDIT.md section 4 for known gaps (validation schema enum drift, dead scripts, UI-only OAuth buttons, no secrets encryption at rest, CSV import not wired, etc.).

S
Description
No description provided
Readme MIT
3.8 MiB
Languages
TypeScript 89.7%
PLpgSQL 6.4%
CSS 1.8%
JavaScript 1.3%
Shell 0.7%
Other 0.1%