[OPIK-8204] [QA] test: cover the alerts CRUD lifecycle with E2E tests

[AndreiCautisanu]

Details

The alerts area had all 8 taxonomy capabilities uncovered — the only alerts coverage was the empty-state screenshot (empty-states.spec.ts E12) and the workspace-role permission matrix, neither of which exercises a list row or the alert editor. Alerts fire webhooks at customer endpoints, so an untested CRUD path is a real gap. This adds tests/alerts/ covering the happy-path spine, the two page objects, and an SDK-seeded fixture.

• Teardown is mandatory rather than tidiness: alerts.project_id is a nullable CHAR(36) with an index and no FK (migration 000060), and ProjectService.delete only touches ProjectDAO, so an alert outlives the project that scoped it and surfaces in the next spec's list — the same hazard automationRulesCleanup documents for rules.
• alerts.test-alert is deliberately left uncovered: "Test connection" POSTs the draft alert to whatever endpoint is in the form, so against example.com the mutation answers failure. A passing test there would only pin an error toast, so the reason is recorded in the taxonomy instead of claiming coverage. Needs a follow-up that stands up a receiver.

Capability	Tier
alerts.list-alerts, alerts.create-alert, alerts.webhook-destination	@t1-smoke
alerts.edit-alert, alerts.event-triggers, alerts.enable-disable, alerts.delete-alert	@t2-cuj

Frontend changes

Four data-testids, all additive — no behavior change:

• alerts-bulk-delete-button — icon-only button with no accessible name and no aria-label; there was no viable stable selector at all.
• alert-name-input, alert-webhook-url-input — these <Label>s render without htmlFor over React-generated ids (:r1o:-form-item), so getByLabel cannot reach either field and the only available name is placeholder copy. Worth noting this is a genuine a11y gap a screen-reader user would hit, not just a test-tooling inconvenience — flagging for a separate fix rather than papering over it here.
• alert-trigger-<eventType> — a trigger's title also renders in the test-webhook panel's accordion, so a text-scoped lookup resolves there and finds none of the config controls.

Change checklist

• User facing
• Documentation update

Issues

• Resolves #
• OPIK-8204

AI-WATERMARK

AI-WATERMARK: yes

• Tools: Claude Code
• Model(s): Claude Opus 5
• Scope: Test authoring (spec, page objects, fixture), the four FE data-testid additions, and the taxonomy update.
• Human verification: Author reviewed the diff. All specs were run locally against a self-built local OSS stack; regression-detection was confirmed by mutation testing (below), not by green runs alone.

Testing

Local OSS stack built from this branch (docker compose --profile opik up -d, frontend image rebuilt so the new testids are served), workspace default.

cd tests_end_to_end/e2e
npx playwright test tests/alerts/ --reporter=list --repeat-each=2   # 8 passed
python3 ../coverage/tag_lint.py --taxonomy ../coverage/taxonomy.yaml --estate ..   # 55 specs, 0 problems
npx playwright test tests/datasets/ --reporter=list --grep "@t1-smoke"   # 2 passed
pre-commit run --files <changed files>   # eslint + typecheck frontend: Passed

Mutation-tested, not just green. Breaking the Status column (accessorFn: () => true) and rebuilding the FE fails exactly the two status-asserting tests on toBeVisible and leaves the other two passing — so these assertions catch a real regression rather than passing vacuously. Verified again after the review refactor to confirm the wider timeouts didn't mask it.

Cold-start behaviour. An early version used the default 15s action timeout and failed reproducibly on a cold backend: the first concurrent burst after idle takes ~18-20s (measured 17.7s/18.7s response times on plain GETs, all 200s — the backend answers, just slowly), so every worker's opening step timed out with the page still on skeleton rows. The readiness waits now allow 30s, the same allowance experiment-detail, online-evaluation and configuration already use. Verified against a deliberately cold stack (docker restart opik-backend-1): 4/4 pass at 14-17s, precisely the window that previously failed. Then 24/24 across repeat runs.

Teardown verified by global-teardown reporting no alerts to sweep on every run — the fixture cleans up rather than leaning on the run-prefix backstop.

Not run: the full suite. Scoped to the new specs plus a sibling directory, since fixtures/index.ts is on every spec's import path; npx playwright test --list confirms all 115 tests across 52 files still enumerate.

Pre-existing, not touched: tests_end_to_end/e2e/tsconfig.json still sets baseUrl, removed in TypeScript 7, so a bare npx tsc --noEmit fails on a clean tree (confirmed by stashing this branch's changes). Typechecking with that line dropped shows 0 errors in the new files. Probably worth its own cleanup ticket.

Documentation

None — internal test coverage only.

🤖 Generated with Claude Code

[CometActions]

Already covered by a test in this PR.

The five frontend files are test affordances only — three data-testids, the shared alertTriggerTestId helper and the bulk-delete button id — with no behaviour change, and the spec this PR ships is what consumes them: tests_end_to_end/e2e/tests/alerts/alerts-crud.spec.ts drives create through the form and re-opens the edit form to assert the webhook URL persisted, edits name/triggers/enabled and round-trips the threshold and window, and deletes one alert then bulk-deletes the rest with an API check that nothing is left. That covers the seven capabilities the taxonomy diff flips to covered, so there is nothing here for a second test to add. One note for review rather than triage: alert-editor.page.ts re-implements the :→- slug inline instead of importing alertTriggerTestId, so the shared helper is shared by comment only.

Not testable yet. alerts.test-alert stays uncovered, and the taxonomy comment in this PR is right about why: Test connection / Test trigger in TestWebhookSection.tsx POST the draft alert to the endpoint in the form, so against example.com only the failure toast is assertable. Recording it as a deferral rather than letting it sit as the one covered: false nobody re-reads — the missing piece is estate-side, not yours.

areas: alerts

Run

Advisory, from the QA test radar. Nothing here blocks this PR, and anything it proposes is a draft for review.

Re-checked after a push on 03 Sep 10:13 UTC — nothing the verdict depends on changed.

[github-actions]

⏱️ pre-commit per-hook timing

Hook	Description	Result	Duration
🌐 typecheck — frontend	Whole-project tsc type check	✅	31.16s
🌐 eslint — frontend	Lint + autofix JS/TS	✅	3.62s
Total (2 ran)			34.78s

⏭️ 42 skipped (no matching files changed)

Hook	Description	Result
🐍 trim trailing whitespace — python sdk	Strip trailing whitespace	⏭️
🐍 fix end of files — python sdk	Ensure files end in a newline	⏭️
🐍 ruff — python sdk	Lint + autofix Python (ruff)	⏭️
🐍 ruff-format — python sdk	Format Python code (ruff)	⏭️
🐍 mypy — python sdk	Static type check	⏭️
🤖 trim trailing whitespace — optimizer	Strip trailing whitespace	⏭️
🤖 fix end of files — optimizer	Ensure files end in a newline	⏭️
🤖 check yaml — optimizer	Validate YAML syntax	⏭️
🤖 check json — optimizer	Validate JSON syntax	⏭️
🤖 check toml — optimizer	Validate TOML syntax	⏭️
🤖 check for added large files — optimizer	Block large files (>1MB)	⏭️
🔐 detect private key — optimizer	Block committed private keys	⏭️
🤖 check for merge conflicts — optimizer	Block merge-conflict markers	⏭️
🤖 check for case conflicts — optimizer	Block case-only name clashes	⏭️
🤖 pyupgrade — optimizer	Modernize Python syntax	⏭️
🤖 ruff — optimizer	Lint + autofix Python (ruff)	⏭️
🤖 ruff-format — optimizer	Format Python code (ruff)	⏭️
🤖 mypy — optimizer	Static type check	⏭️
📓 nbstripout — optimizer notebooks	Strip notebook output	⏭️
📝 markdownlint — optimizer	Lint Markdown	⏭️
🔤 codespell — optimizer	Fix common misspellings	⏭️
📊 radon cc — optimizer	Cyclomatic-complexity gate	⏭️
📊 radon raw — optimizer	Raw size metrics gate	⏭️
📊 xenon — optimizer	Fail on complexity thresholds	⏭️
📊 lizard — optimizer	Cyclomatic-complexity gate	⏭️
🧹 vulture — optimizer	Find dead code	⏭️
🛡️ trim trailing whitespace — guardrails	Strip trailing whitespace	⏭️
🛡️ fix end of files — guardrails	Ensure files end in a newline	⏭️
🛡️ ruff — guardrails	Lint + autofix Python (ruff)	⏭️
🛡️ ruff-format — guardrails	Format Python code (ruff)	⏭️
🛡️ mypy — guardrails	Static type check	⏭️
⚓ helm-docs	Regenerate Helm chart README	⏭️
block non-public FE plugins	Block non-public FE plugins	⏭️
☕ spotless — java backend	Format Java code	⏭️
🧪 pre-commit wrapper smoke tests	Self-test the wrapper scripts	⏭️
🧪 rebaseline script tests	Self-test the changelog re-baseline script	⏭️
📘 eslint — typescript sdk	Lint + autofix JS/TS	⏭️
📘 typecheck — typescript sdk	Whole-project tsc type check	⏭️
⚙️ actionlint — github workflows	Lint GitHub Actions workflows	⏭️
🐳 hadolint — dockerfiles	Lint Dockerfiles	⏭️
🌈 zizmor — github workflows security	Security-scan GitHub Actions workflows	⏭️
🛡️ semgrep — java backend sql	Block SQL injection-prone string formatting	⏭️

[aadereiko]

LGTM!