[wnykuang]

Summary

Fixes #290 — Playwright CI tests were broken and the workflow was disabled.

Root causes and fixes:

• Node 17 → 18: package.json requires "engines": {"node": "^18"}, but the workflow specified Node 17. Updated to Node 18.
• Ubuntu 22.04 → 24.04: manifest.json targets Ubuntu 24.04 binaries (Ruby, MongoDB, OpenStudio). Updated the runner to match.
• Electron system dependencies: Electron 21 requires GTK3, NSS, ALSA, and other system libraries that aren't pre-installed on Ubuntu 24.04 runners. Added explicit apt-get install step with the correct t64 package names for Noble.
• Electron sandbox: Electron's Chrome sandbox fails on CI runners. Added --no-sandbox flag when process.env.CI is set (in playwright/App.ts).
• Workflow triggers: Scoped from branches: "*" to develop + tags + PRs, matching the build_pat.yaml pattern.

Serial server tests skipped in CI:

The serial test suite includes integration tests that start/stop the actual OpenStudio server (MongoDB + Ruby + EnergyPlus). These require a fully configured runtime stack and time out in GitHub Actions — the stopServerTests were already skipped in CI for this reason. Extended the same !process.env.CI guard to startServerTests, serverPageTests, and runPageTests, which all depend on a running server. The saveProjectTests still runs in CI since it doesn't need the server.

Changes

• .github/workflows/playwright.yml.disable → .github/workflows/playwright.yml (re-enabled)
• playwright/App.ts — pass --no-sandbox to Electron in CI
• playwright/tests/serial/serial.list.ts — skip server-dependent tests in CI

Test plan

• All 188 parallel tests pass in CI
• Serial saveProjectTests passes in CI (1 passed)
• Server-dependent serial tests skip cleanly in CI
• All 22 serial tests still register when running locally
• Verified in fork: https://github.com/wnykuang/OpenStudio-PAT/actions/runs/21775205736

[Copilot]

Warning

Copilot couldn't run its full agentic review because it didn't start before the timeout. Make sure your repository has a runner available, or add a copilot-code-review.yml file specifying one with the runs-on attribute. See the docs for more details.

Copilot review overview

1 open finding

• Using --no-sandbox meaningfully reduces Chromium/Electron process isolation. Since the workflow… · New

What changed in this PR

Re-enables the Playwright CI workflow and updates it to run reliably on the current project/runtime requirements (Node 18, Ubuntu 24.04, Electron sandbox/deps), while skipping server-dependent serial tests in CI.

Changes:

• Updated Playwright serial test list to skip server-dependent suites when running in CI.
• Updated Playwright Electron launch args to disable the sandbox in CI.
• Updated the Playwright GitHub Actions workflow (runner, Node version, deps install, caching, triggers, concurrency).

File	Description
playwright/​tests/​serial/​serial.list.ts	Skips server-dependent serial tests under CI to prevent timeouts.
playwright/​App.ts	Adds --no-sandbox to Electron launch args in CI to avoid sandbox failures.
.github/​workflows/​playwright.yml.disable	Updates workflow configuration for Node 18/Ubuntu 24.04 and installs Electron system dependencies.

🧠 Review effort: Lite

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.