perf(release): build Vercel deployments early and go live after publish

[t3dotgg]

The last two release jobs wait on Vercel builds. In the last nightly (run), Deploy hosted web app took 177s (141s of it vercel deploy) and Deploy marketing site took 76s, both after the GitHub Release published. The Discord announcement waited on the web deploy. But the builds only need the release commit and version, which preflight knows 20 minutes earlier. The marketing site reads release data in the browser, not at build time.

Each deploy now has two jobs:

• build_web and build_marketing start right after preflight, in parallel with the desktop builds. They run vercel deploy --prod --skip-domain, which builds a production deployment without moving the custom domains, and pass the deployment URL on as a job output.
• deploy_web and deploy_marketing still wait for the GitHub Release. They only run vercel alias set (web channel and router domains, same logic as before) or vercel promote (marketing, which Vercel documents as the way to finish a --skip-domain deployment). Each runs on a 2 vCPU runner with no dependency install.

The custom domains (app.t3.codes, the nightly and latest channel domains, t3.codes) do not change earlier than before. --skip-domain only skips custom domains, though: each project's own *.vercel.app production hostname moves at build time, before the release. For the web app that hostname (t3code-app-pinglabs.vercel.app) is behind Vercel SSO and redirects the public to a login page, so users still only receive a client after its npm package exists, and the order in docs/operations/release.md holds for them. The docs now say this. The marketing site's vercel.app hostname is public, but the site has no version-ordering rule and reads release data in the browser. If the release fails, the custom domains stay on the previous deployment, the same as when an alias step fails today. The web deploy used --skip-domain already. The marketing deploy used to go live directly and now goes through promote.

Expect the deploy step after the release to drop from about 177s to about 25s, and the Discord announcement to post about 2.5 minutes sooner. I did not run these with Vercel credentials. actionlint and format pass, and the pinned vercel@53.1.1 help confirms both --skip-domain and promote --yes. A nightly will confirm.

Created with Claude Opus 5.5 in Claude Code, running inside T3 Code.

🤖 Generated with Claude Code

[macroscopeapp]

Approvability

Verdict: Not approved

Macroscope's review found this PR not approvable — This PR changes production Vercel deployment timing and promotion semantics by creating staged production deployments before release publication, including movement of project-owned production hostnames. The multi-job release orchestration and altered failure/publication behavior are significant enough to require human review.

No code changes detected at 9e2ffa4. Prior analysis still applies.

^{You can add or adjust custom eligibility rules. Learn more.}

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	13.5 KiB	13.5 KiB	−3 B (−0.0%)	15.1 KiB	✅
Codex	Thread snapshot wire	7.1 KiB	7.1 KiB	+3 B (+0.0%)	7.3 KiB	✅
Codex	Live turn WebSocket wire	6.4 KiB	6.4 KiB	−6 B (−0.1%)	7.8 KiB	✅
Codex	Live turn WebSocket decoded	56.2 KiB	56.2 KiB	0 B (0.0%)	66.4 KiB	✅
Codex	Live turn messages	9	9	0 (0.0%)	21	✅
Claude	Total thread wire	13.5 KiB	13.5 KiB	−2 B (−0.0%)	15.1 KiB	✅
Claude	Thread snapshot wire	7.1 KiB	7.1 KiB	−2 B (−0.0%)	7.3 KiB	✅
Claude	Live turn WebSocket wire	6.4 KiB	6.4 KiB	0 B (0.0%)	7.8 KiB	✅
Claude	Live turn WebSocket decoded	57.0 KiB	57.0 KiB	0 B (0.0%)	66.4 KiB	✅
Claude	Live turn messages	9	9	0 (0.0%)	21	✅

Baseline: c57a04b · PR result: 9e2ffa4 · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 114.0 KiB
• Claude decoded thread snapshot: 114.7 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.

[coderabbitai]

Navigate logical layers of code changes, visualize relationships, and explore their blast radius.

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Repository: pingdotgg/t3code/.coderabbit.yaml

Review profile: CHILL

Plan: Team

Run ID: 1427668d-657b-4e9e-af06-1e9feac67935

📥 Commits

Reviewing files that changed from the base of the PR and between 7aedf26 and 9e2ffa4.

📒 Files selected for processing (1)

• .github/workflows/release.yml

Included review availability: This review used your included allowance. Your plan provides up to 10 included reviews per hour; 2 remain after this review.

---

📝 Walkthrough

Walkthrough

The release workflow stages production web deployments before release publication. After publication, it assigns the web deployment’s channel alias. For stable releases, it also assigns the router alias when applicable. Nightly releases stage a marketing deployment before publication and promote it afterward.

Changes

Release deployments

Layer / File(s)	Summary
Stage the web deployment .github/workflows/release.yml, docs/operations/release.md	The workflow stages a production web deployment with --skip-domain and exports its URL. The operations guide describes the staging timing and production hostname behavior.
Assign web deployment aliases .github/workflows/release.yml	After the GitHub Release succeeds, the workflow assigns the channel alias and, for stable releases, the router alias when it is configured and differs from the channel domain.
Stage the nightly marketing deployment .github/workflows/release.yml, docs/operations/release.md	For nightly releases, the workflow stages a production marketing deployment with --skip-domain and exports its URL and project ID. The operations guide describes staging during desktop builds.
Promote the nightly marketing deployment .github/workflows/release.yml, docs/operations/release.md	After the GitHub Release succeeds, the workflow promotes the staged marketing deployment. The operations guide describes promotion after publication.

Priority: ➖ Normal

Estimated code review effort: 2 (Simple) | ~15 minutes

Change: Refactor

Sequence Diagram(s)

sequenceDiagram
  participant ReleaseWorkflow
  participant Vercel
  participant GitHubRelease
  participant WebDomains
  ReleaseWorkflow->>Vercel: Stage web deployment with --skip-domain
  ReleaseWorkflow->>Vercel: Stage nightly marketing deployment with --skip-domain
  GitHubRelease-->>ReleaseWorkflow: Release succeeds
  ReleaseWorkflow->>Vercel: Alias staged web deployment to channel domain
  Vercel->>WebDomains: Assign channel alias
  ReleaseWorkflow->>Vercel: Assign stable router alias when applicable
  ReleaseWorkflow->>Vercel: Promote staged nightly marketing deployment

Loading

Suggested reviewers: juliusmarminge

Merge Risk: ⚪ Minimal · up to 9e2ff

The workflow stages deployments earlier while retaining post-release custom-domain publication. No actionable merge-blocking risk was identified; credentialed deployment behavior still requires normal operational validation.

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Description check	⚠️ Warning	The description explains the problem, implementation, expected impact, verification performed, and agent usage. It does not provide the required Scope and approval section or link triaged issue discus…	Add a Scope and approval section with the relevant triaged issue or maintainer approval, including the approval comment. State why the workflow and deployment behavior changes remain within the approved scope.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly and concisely describes the main change: moving Vercel builds earlier and promoting deployments after publication.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check. Docstring coverage is scoped to functions touched by this diff. Analyzed 0 functions across 0…
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

Full details: Description check

Explanation

The description explains the problem, implementation, expected impact, verification performed, and agent usage. It does not provide the required Scope and approval section or link triaged issue discussion and maintainer approval for this broader release workflow change.

• Fix all pre-merge checks with AI

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Commit to this branch
• Create a new PR

---

_{Comment @coderabbitai help to get the list of available commands.}