Add Jest + Playwright BrowserStack SDK sample

[saraswati-padmawar]

Populates this customer-facing sample repo (previously skeleton: README + CODEOWNERS + Semgrep only) with a working Jest + Playwright sample driven by the BrowserStack Node SDK.

What's included

• tests/bstack_sample.test.js — public bstackdemo.com add-to-cart scenario
• tests/bstack_local_test.test.js — BrowserStack Local scenario (bs-local.com:45454)
• browserstack.yml — credentials, 3 platforms (Win/Chrome, macOS/WebKit, Win/Firefox), Local toggle, framework: playwright, testReporting: true
• package.json — sample-test / sample-local-test scripts via browserstack-node-sdk jest
• jest.config.js / jest.local.config.js, .gitignore, README

References

• Follows the standardized Automate sample template (Confluence: SDK Sample Repo Generator, R1–R4)
• SDK run command per the node-agent framework-matrix.md (browserstack-node-sdk jest)
• framework: playwright mirrors the existing codecept-js-playwright-browserstack precedent
• Structure mirrors node-js-playwright-browserstack

Verified on BrowserStack

Project: BrowserStack Samples.

The public scenario is fully verified. The Local scenario could not be confirmed in the local environment used for verification — the SDK's BrowserStack Local tunnel was not connected by the time the parallel Jest workers called chromium.launch() ([browserstack.local] is set to true but local testing through BrowserStack is not connected). The same Local scenario passes in the sibling cucumber-js and nunit samples with identical config, so this is a tunnel-startup timing issue in that environment, not a defect in the sample's test code or browserstack.yml. The included sanity-workflow.yml runs the Local step on a clean GitHub runner — please confirm it green there before merge.

Note from verification: the sample must depend on @playwright/test (not playwright) for the SDK to detect & route Playwright — reflected in this PR.

🤖 Generated with Claude Code

[Rohannagariya1]

LGTM ✅ — public scenario verified and the sample is clean. Approving.

One thing worth confirming before merge (non-blocking on the code itself):

• The Local scenario is unverified due to the SDK tunnel not being connected when the parallel Jest workers call chromium.launch(). The "passes in the cucumber/nunit siblings" argument doesn't fully carry over since those aren't multi-worker Jest — the tunnel-startup vs parallel-worker race can reproduce on the runner too. Please confirm the sanity workflow is green on the Local step before merging; if it flakes, maxWorkers: 1 in jest.local.config.js would serialize the workers and sidestep it.

Optional/minor:

• latest deps + postinstall: npm update + gitignored lockfile = non-reproducible installs (matches BrowserStack sample convention, fine to keep).

[saraswati-padmawar]

Claude Code PR Review

PR: #2 • Head: 8e7c834 • Reviewers: stack:code-reviewer

Summary

Populates this previously-skeleton sample repo with a working Jest + Playwright sample driven by the BrowserStack Node SDK. The public bstackdemo scenario is verified green across 3 platforms; the BrowserStack Local scenario is pending CI confirmation.

Review Table

Priority	Category	Check	Status	Notes
High	Security	No hardcoded secrets or credentials	Pass	Placeholders only; org secrets in workflow
High	Security	Authentication/authorization checks present	N/A	Sample test project
High	Security	Input validation and sanitization	N/A	—
High	Security	No IDOR — resource ownership validated	N/A	—
High	Security	No SQL injection (parameterized queries)	N/A	No DB
High	Correctness	Logic is correct, handles edge cases	Pass	testMatch patterns correct; public run verified green
High	Correctness	Error handling is explicit, no swallowed exceptions	Pass	beforeAll/afterAll manage browser lifecycle
High	Correctness	No race conditions or concurrency issues	Pass	Per-suite browser; SDK fans platforms into separate workers
Medium	Testing	New code has corresponding tests	Pass	The sample is the test
Medium	Testing	Error paths and edge cases tested	N/A	Sample scope
Medium	Testing	Existing tests still pass (no regressions)	Pass	Fresh repo
Medium	Performance	No N+1 queries or unbounded data fetching	N/A	—
Medium	Performance	Long-running tasks use background jobs	N/A	—
Medium	Quality	Follows existing codebase patterns	Pass	Mirrors node-js-playwright-browserstack
Medium	Quality	Changes are focused (single concern)	Pass	—
Low	Quality	Meaningful names, no dead code	Pass	except main: index.js (below)
Low	Quality	Comments explain why, not what	Pass	—
Low	Quality	No unnecessary dependencies added	Pass	@playwright/test required for SDK PW detection

Findings

README.md Medium — Local test not runnable as documented (no server on :45454; title assertion coupled to CI harness) Resolved — fixed in 8e7c834: README now documents serving a matching local page on 45454 and adapting to your own app.

• File: tests/bstack_local_test.test.js:31 + browserstack.yml

• Severity: Medium

• Reviewer: stack:code-reviewer

• Issue: the BrowserStack Local scenario is not yet verified end-to-end (SDK Local tunnel-startup timing in the verification environment); the public scenario is verified green.

• Suggestion: treat a green sanity-workflow.yml Local step on the runner as a merge gate before relying on the Local path.

• File: package.json:5

• Severity: Low

• Reviewer: stack:code-reviewer

• Issue: "main": "index.js" references a file that does not exist (leftover from npm init).

• Suggestion: remove the main field (and the unused keywords/author boilerplate).

• File: jest.config.js:7 / jest.local.config.js:6

• Severity: Low

• Reviewer: stack:code-reviewer

• Issue: testTimeout: 90s includes remote session startup + nav + assertions; may be tight under session queueing on a busy account.

• Suggestion: consider 120s for headroom.

• File: .github/workflows/sanity-workflow.yml:19

• Severity: Low

• Reviewer: stack:code-reviewer

• Issue: CI matrix only exercises Node 20 though README advertises 18/20/22.

• Suggestion: add 18/22 to the matrix, or keep the "verified on Node 20" caveat prominent.

---

Verdict: PASS — no High/Critical issues; the public scenario is verified green, testMatch and SDK-detection logic are correct, and the workflow has no injection risk. Confirm the Local step on CI before merge.

Review run via the BrowserStack AI harness stack:pr-review methodology (harness not installed in the run environment → independent stack:code-reviewer fallback).