[Brooooooklyn]

The platform was merged into this repo under platform/ but has never been deployed from here. This PR fixes the remaining repo-side defects that block a deploy: the platform CI job order, and the deploy docs. (Main already carries the vp run build step in the deploy workflows and the kitchen-sink Better Auth migration, so those parts were dropped from this PR after a rebase.)

platform-ci.yml    fmt → lint → build → platform build → [lint:platform] → [check:platform] → [check:sdk]
                                                          moved after builds   was: check          new
platform docs      README.md, meta/operations.md describe the monorepo workflows

What changed

1. platform-ci.yml: run lint:platform after the builds; use check:platform. lint:platform reads platform/packages/dashboard/tsconfig.json, which extends the generated .void/tsconfig.json from the dashboard build. The root check also typechecks the nine playgrounds, whose tsconfigs extend generated files this job never builds. Both deploy workflows call this job via workflow_call, so it must pass on a fresh checkout. The lint:platform / check:platform hunk is identical to the one on feat(email): void/email — outbound, inbound, platform proxy, and own-account setup #316.

   check:sdk (new root script, vp run --filter './packages/**' --parallel typecheck) runs after check:platform. The deploy workflows run only this job, the dashboard and the workers bundle the SDK, and vp pack does not typecheck, so without it a ref with SDK type errors could deploy. It takes ~2 s and needs no playground builds. The playgrounds stay with ci.yml.

2. platform/README.md, platform/meta/operations.md: describe the monorepo workflows (manual dispatch, the VOID_PLATFORM_DEPLOY_ENABLED gate, step order, platform/package.json as the deploy orchestrator, the SDK build prerequisite) instead of the old push-to-prod / push-to-main model.

Verified

• actionlint (with the pre-existing runner-label ignores) and vp fmt --check pass.
• vp run lint:platform fails on a clean tree and passes after the two builds; vp run check:platform and vp run check:sdk pass without the playground builds.
• wrangler deploy --dry-run passes for all five platform workers (staging and production) from the monorepo.

Already done outside this PR

• Repo config: VOID_PLATFORM_DEPLOY_ENABLED=true; repo secrets CF_API_TOKEN (User API token) and CF_ACCOUNT_ID; environment Prod with DATABASE_URL; environment Staging with STAGING_DATABASE_URL, CF_ACCESS_CLIENT_ID_STAGING, CF_ACCESS_CLIENT_SECRET_STAGING. The environments have no protection rules yet.

Not in this PR

• Cutover: disable deploy-staging.yml, deploy.yml, and smoke-test.yml in voidzero-dev/void-platform before the first dispatch here. The old repo still auto-deploys staging on push to its main.
• schema-compat stays if: false until a production ref is chosen. Suggested: a platform-prod branch that platform-deploy.yml moves after a successful deploy; the existing prod branch in this repo is an unrelated SDK branch from June.
• SDK follow-up (product call): any existing managed app with auth enabled hits the drift check after upgrading past 754d9a7, and a plain generated migration fails on its D1 because the runtime already created the tables. kitchen-sink is covered by main's IF NOT EXISTS migration; the general case needs an SDK decision.

🤖 Generated with Claude Code

https://claude.ai/code/session_01WNXbwieSdsSErZwQ7HWRwG

---

Note

Low Risk
Changes are limited to CI step ordering, a new npm script, and documentation; no runtime or deploy logic changes.

Overview
Fixes platform CI so reusable deploy workflows can pass on a fresh checkout, and updates platform runbooks to match the merged monorepo’s manual deploy model.

CI (platform-ci.yml + root check:sdk): lint:platform now runs after SDK and platform package builds, because it depends on the dashboard’s generated .void/tsconfig.json. The job no longer runs root vp run check (which typechecks playgrounds that this job never builds); it runs check:platform and the new check:sdk script instead.

Docs (platform/README.md, platform/meta/operations.md): Deploy triggers are documented as manual platform-deploy* workflows gated by VOID_PLATFORM_DEPLOY_ENABLED, with platform/package.json (or vp run --filter @voidcloud/platform …) as the deploy orchestrator and an explicit SDK vp run build prerequisite before dashboard deploy.

^{Reviewed by Cursor Bugbot for commit cb7ec08. Bugbot is set up for automated code reviews on this repo. Configure here.}

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-09-13T15:43:55.592397Z	cb7ec08	New commits

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[cursor]

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: serverGenReqId_5fd52083-dc59-40e2-b4fc-b4ece42c4fca)

[cursor]

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: serverGenReqId_f100fae5-9669-47e4-8c1d-944fe421a568)

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 2b146f84a1

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[cursor]

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: serverGenReqId_d8b9ceb7-835f-44f8-9ecf-3a1b886a4108)