[nick-potts]

Summary

Implements spec #986: the first stable tree. Cloud authors the Deploy Intent, and the CLI operates a Cluster over SSH. #987–#992 and #994 already landed on main (#994 in #1059). This branch adds the rest.

 dashboard/
-├── routes/ waitlist, org logs/deployments, docs/home/pricing   # #993
+├── routes/_public/index.tsx         # stub lander; /auth unchanged
-├── drizzle/ 12 migrations            # #995
+├── drizzle/<ts>_baseline/           # one fresh baseline, no waitlist
-├── github-ingestion.postgres-test-harness.ts  # promise harness, #995
+├── self-host/compose.yml, .env.example        # web·worker·Inngest·Redis·Postgres, #996
 └── server/config.server.ts          # names every missing var
 core/
-└── PLOYZ_SSH_CONTROL_PERSIST, PLOYZ_HEALTH_MONITOR_PERIOD,
-    PLOYZ_RELEASE_DIR, PLOYZ_UPGRADE_RELEASE_DIR, --release-dir  # #1025
-docs/{adr,audits,design}/  evidence/  Dockerfile  .dockerignore # #997
+docs/agents/domain.md                # only doc kept

Release flow (#996):

tag vX.Y.Z
  cloud job → ghcr.io/getployz/ployz-cloud:vX.Y.Z   (exact tag, no moving tags)
  publish   → attach ployz-cloud-compose.yml + ployz-cloud.env.example
self-host
  docker compose run --rm web npm run db:migrate
  docker compose up        # no Polar vars = billing off, unlimited Organizations

Missing config now fails like this:

ConfigError(SchemaError(Expected string at ["DATABASE_URL"]
                        Expected string at ["APP_URL"] … one line per var))

Evidence

• Before: A missing variable gave a ConfigError that didn't say which one. There was no self-host compose, and there were 12 migrations plus two Postgres test harnesses.
  After:
  • The config test boots with an empty env and checks that all 12 required names appear in the error.
  • docker compose up on a fresh database with no Polar variables: web returns 200 on / and /api/auth/get-session, the worker is healthy, and the baseline applies.
  • baseline.postgres.test passes on an empty database. The dashboard's 975 tests passed after Squash Cloud schema to one fresh Postgres baseline #995, including 30 Postgres test files.
  • Core: cargo clippy and nextest pass on the merged tree. actionlint and scripts/test-ci.py pass.
  • Verify the cleaned 1.0 tree end to end #998: 1522 nextest tests, the SDK contract, installer and CI tests, and the dashboard's 989 tests all pass on the merged tree.
  • Four-axis review: Standards, Spec, Maintainability and Simplicity all PASS after 4 rounds.
  • Fast CI is green at 48efe0e, covering Rust, SDK types, contracts, macOS CLI, release builds and smoke tests, and Cloud static, test and build.

Merge Danger

Door: one-way

The migration history is squashed into one baseline, so the existing Railway Postgres has to be dropped (it's disposable per #986). ADRs, audits, design notes and evidence/ are deleted; they remain in git history.

Blast Radius: wide

• Open PRs: any open PR that adds drizzle migrations must regenerate them on top of the new baseline, including Hosted DNS, Cluster Domain and wildcard certificates #1061 (Hosted DNS) and Deployment mode: deployments as a mode of the canvas #1056 (deployment mode). PRs that link to the deleted docs need rebasing.
• Image: the Cloud image is linux/amd64 only. The draft release run pushes ployz-cloud:<tag> after tests pass, so the tag is pullable before Publish. RELEASE.md says so.
• CI time: Cloud's build job now waits for static and test.
• Releases: a release run now also runs Cloud's static checks and tests.
• Not in scope: Hosted DNS code still says dns.uncloud.run, and PLOYZ_INGRESS_VERIFY_PORT remains until Remove Hosted DNS from the daemon, SDK and CLI #1024.

Closes #986
Closes #993
Closes #995
Closes #996
Closes #997
Closes #998
Closes #1025

[chatgpt-codex-connector]

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.
To continue using code reviews, add credits to your account and enable them for code reviews in your settings.