experiment: test 12x runners on CI workflow

[lawrencecchen]

Testing warp-macos-15-arm64-12x (12 vCPU, 44 GB RAM, 270 GB SSD, $0.16/min) vs current 6x (6 vCPU, 22 GB RAM, 120 GB SSD, $0.08/min) on the CI workflow.

Baseline on 6x (avg of 5 recent runs):

• tests job: 6m41s ($0.53/run)
• tests-build-and-lag job: 4m34s ($0.37/run)

Break-even point: if 12x cuts time by >50%, it's cheaper per run despite 2x $/min. Xcode builds are heavily parallelizable so this is plausible.

What to compare after this runs:

• Job durations on 12x vs baseline above
• WarpBuild observability: CPU/disk/memory utilization (should drop significantly with 270 GB disk and 12 cores)

Do not merge. Experiment only.

---

Summary by cubic

Run CI tests and tests-build-and-lag on warp-macos-15-arm64-12x to measure speed vs cost versus 6x; break-even is ≥50% faster given 2x price/min. Updated the CI guard to allow 6x or 12x while keeping the fork billing guard.

• Refactors
  • Switched both jobs to runs-on: warp-macos-15-arm64-12x.
  • Updated tests/test_ci_self_hosted_guard.sh to accept warp-macos-15-arm64-(6|12)x and simplified error messages.

^{Written for commit e24b66b. Summary will update on new commits.}

Summary by CodeRabbit

• Chores
  • Updated CI infrastructure with improved runner configurations for enhanced test execution performance and compatibility.

[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.

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
cmux	Ready	Preview, Comment	Mar 18, 2026 4:03am

[greptile-apps]

Your free trial has ended. If you'd like to continue receiving code reviews, you can add a payment method here.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 0ba80fc6-7ce1-4d7f-bb8a-a4ff40c84f21

📥 Commits

Reviewing files that changed from the base of the PR and between 3bca43d and e24b66b.

📒 Files selected for processing (2)

• .github/workflows/ci.yml
• tests/test_ci_self_hosted_guard.sh

---

📝 Walkthrough

Walkthrough

The changes upgrade the CI workflow to use a more capable WarpBuild macOS ARM64 runner (12x instead of 6x) and update the corresponding validation script to accept both runner variants (6x and 12x).

Changes

Cohort / File(s)	Summary
CI Workflow Configuration .github/workflows/ci.yml	Updated runner specifications from warp-macos-15-arm64-6x to warp-macos-15-arm64-12x in tests and tests-build-and-lag jobs.
CI Guard Validation tests/test_ci_self_hosted_guard.sh	Expanded regex pattern to accept both 6x and 12x runner variants; updated failure messages to reference generic warp-macos runner instead of specific 15-arm64-6x variant.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~3 minutes

Possibly related PRs

• Set up full test suite in CI and Xcode Cloud #447: Modifies the self-hosted CI guard script and CI workflow's tests job for runner specification updates
• Switch CI and Build GhosttyKit to Depot macOS runners #710: Updates CI guard and workflow runner selection to accept additional macOS self-hosted runner variants
• Migrate CI/CD to WarpBuild, consolidate test jobs #1501: Modifies CI runner labels and self-hosted guard script to reference WarpBuild macOS runners

Poem

🐰 The runners now grow stronger, twelve times might,
From six to double, we ascend new height,
The guard script nods, accepting both with grace,
Warp builds leap faster in the CI race! ⚡

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title accurately summarizes the main change: testing 12x runners on the CI workflow, which matches the core objective of experimenting with a more capable WarpBuild macOS ARM64 runner.
Description check	✅ Passed	The description includes most required sections: summary (what changed and why), testing approach, and notes. However, it lacks explicit testing steps, a demo video section is N/A, and the checklist items are not addressed.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch experiment-12x-ci

📝 Coding Plan

• Generate coding plan for human review comments

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

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

[cubic-dev-ai]

No issues found across 2 files

[lawrencecchen]

Experiment complete. 12x runners only 7-15% faster but 2x cost/min, not worth it. Staying on 6x.