A desktop client reading /api/software had no way to learn that a title costs money. There was nothing in the response to say so, no way to sign in, and no way to be told "you do not own this" — so a store with paid titles could only hand the client a 403 at download time and let it guess why. The fix belongs here rather than in the client. A client serves more than one store, so anything it knows about a particular one has to arrive from that store's own API; a rule compiled into the client is a rule that breaks every other catalog it reads. Three seams, each following the storage adapter's shape — documented contract, default that is byte for byte the old behaviour, one config key to replace it: - **access policy** — visible_software_scope / access_for / authorize_download. Every catalog entry now carries an `access` block (gated, entitled, price, purchaseUrl, webUrl) and both /api/download and /file/* ask before serving. The vocabulary is deliberately generic: a word from one host's domain would make every client that reads it specific to that host. - **client sign-in** — the device authorization grant (RFC 8628), over the host's own user model. The approval page stays the host's, because approving needs a session and HTML. Tokens are ApplicationTokens with a `catalog` scope, so publishing and reading stay separable. - **service descriptor** — GET /api/service says what this deployment is and whether it has a sign-in at all, which is how a client stops guessing. With no policy and no subject class configured — every deployment today — the API is unchanged: /api/auth/* answers 404, /api/service reports auth: null, and the 187 pre-existing examples pass untouched. A policy that raises is treated as a refusal, not permission. An artifact served because the gatekeeper crashed is the one failure mode this must not have, so a broken policy empties the catalog and denies the download. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Teletype Games
Monorepo for the Teletype Games portal: a Vue 3 frontend, a Rails 8 API host app, and the reusable WarpEngine software-catalog engine.
Project structure
apps/
frontend/ # Vue 3 + Vite + TypeScript + Tailwind SPA
api/ # Rails 8 host app: TTG-specific API + ActiveAdmin shell
libs/
ruby/warp_engine/ # WarpEngine: mountable Rails engine (catalog, updater, admin resources)
The API is split in two layers:
- WarpEngine (
libs/ruby/warp_engine) owns the software catalog: models (softwares, releases, release assets, images, platform links, download stats), the CI-callable/build/*publishing endpoints, the public read-only JSON API (/api/software*,/api/builds*,/api/image,/api/download,/file/*) and the catalog ActiveAdmin resources. See its README. - The host app (
apps/api) owns everything TTG-specific: members, events, the store registry, wiki proxy, RSS feeds, Devise/ActiveAdmin authentication, theming and assets. It consumes WarpEngine as a path gem and mounts it at/.
The store registry
GET /api/stores lists the stores a client can install from. The desktop
graphical client reads it on first run, which is why it is public: a client has
nobody to log in as.
[
{ "name": "Teletype Games", "catalogUrl": "https://teletypegames.org", "storeRepositoryUrl": null }
]
A Store row has two required fields — name and catalog_url — plus an
optional store_repository_url, maintained from ActiveAdmin ▸ 🛒 Stores;
db/seeds.rb creates our own.
A store does not need a repository. The store engine's own defaults already
cover the host-to-asset mapping, the install modes, the platforms and the
behaviour; what they cannot know is identity — a slug, a name and a catalog — and
that is what this row carries. With no repository the client derives the slug from
the catalog host, writes a small config and installs. Given one, it reads that
repository's config.json and points it at catalog_url, and that file remains the
authority on how the store behaves; a repository without a config file is treated
as no repository at all.
Every WarpEngine API response also carries a WarpEngine-Version header — the registry
above is the host's own endpoint and does not, because it is not part of the engine.
This lives in the host app on purpose, not in WarpEngine. The engine serves one catalog and has no business knowing which stores exist for it; who ships a store for a catalog is a property of the site.
| Model | apps/api/app/models/store.rb |
| Endpoint | apps/api/app/controllers/api/stores_controller.rb |
| Admin | apps/api/app/admin/stores.rb |
| Client | warp-engine-client |
Development environment
docker compose up -d
| Service | URL |
|---|---|
| Frontend | http://${WEBAPP_DOMAIN} |
| API | http://${WEBAPP_DOMAIN}/api |
| Admin | http://${WEBAPP_DOMAIN}/admin |
| API docs | http://${WEBAPP_DOMAIN}/api/swagger |
The api image is built from the repo root (so the libs/ path gems are visible
during bundle install) and mounts ./libs at runtime. After changing the
compose file or the Gemfile, rebuild with docker compose up -d --build api.
Testing
make api-test
Runs the host suite (apps/api, softwares_test DB), the WarpEngine suite
(dummy app, warp_engine_test DB) and a production-mode zeitwerk:check.
Individually:
docker exec -e RAILS_ENV=test api bundle exec rspec # host
docker exec -w /libs/ruby/warp_engine -e RAILS_ENV=test api bundle exec rspec # engine
JSON contract baselines for /api/software and /api/builds live in
apps/api/spec/snapshots/ — diff against them after refactors.
Linting
Backend (RuboCop)
docker compose run --rm --no-deps api bundle exec rubocop # check
docker compose run --rm --no-deps api bundle exec rubocop -A # autofix
Config: apps/api/.rubocop.yml (rubocop-rails-omakase preset)
Frontend (ESLint)
docker compose run --rm --no-deps frontend npm run lint # check
docker compose run --rm --no-deps frontend npm run lint:fix # autofix
Config: apps/frontend/eslint.config.js (ESLint 9 flat config, Vue + TypeScript)