Harden E2E seed loading for test_cases schema drift and add preflight seed integrity check

[Copilot]

E2E startup was seeding requirements but dropping seeded test cases/links when test_cases.auto_registered is present and non-nullable, causing deterministic Traceability Matrix failures (TC-004 missing, zeroed coverage). This PR aligns seed behavior with live schema and fails fast when seed integrity is incomplete.

• Diagnostic findings

  • Confirmed root cause (Hypothesis 1): frontend/tests/e2e/fixtures/seed.sql inserted test_cases without auto_registered, triggering NOT NULL violations and cascading FK failures for requirement_test_case_links/link_suggestions.
  • Not observed (Hypothesis 2): E2E compose startup does not invoke backend/app/db/sample_data.py.
  • Not observed (Hypothesis 3): no TRUNCATE/DELETE FROM test_cases hooks in frontend/backend test setup.
  • Why [Test] E2E CRUD specs pollute seeded TC-00X rows, cause flaky Traceability Matrix failures #308 didn’t fix this: [Test] E2E CRUD specs pollute seeded TC-00X rows, cause flaky Traceability Matrix failures #308 addressed CRUD pollution; this failure path was seed insertion/schema mismatch.

• Seed compatibility + startup fail-fast

  • Updated seed.sql to:
    • enable ON_ERROR_STOP
    • seed external_id for TC-001..TC-005
    • conditionally include auto_registered=false when the column exists.
  • Updated backend seed execution (backend/entrypoint.sh) to run psql -v ON_ERROR_STOP=1 so partial seed failures stop container startup.

• E2E smoke check (pre-test guardrail)

  • Added Playwright globalSetup (frontend/tests/e2e/global-setup.ts) and wired it in frontend/playwright.config.ts.
  • Preflight assertions:
    • requirements >= 5
    • test_cases >= 5
    • links >= 3
    • TC-001..TC-005 present by external_id
  • Throws explicit errors naming missing seed entities to prevent latent data issues from surfacing as downstream UI failures.

-- seed.sql (schema-aware test_case insert path)
IF EXISTS (
  SELECT 1
  FROM information_schema.columns
  WHERE table_schema = current_schema()
    AND table_name = 'test_cases'
    AND column_name = 'auto_registered'
) THEN
  INSERT INTO test_cases (..., external_id, ..., auto_registered, ...)
  VALUES (..., 'TC-004', ..., false, ...);
END IF;

Warning

Firewall rules blocked me from connecting to one or more addresses (expand for details)

I tried to connect to the following addresses, but was blocked by firewall rules:

• 127.0.0.11
  • Triggering command: REDACTED, pid is -1 (packet block)
• dl.google.com
  • Triggering command: /usr/lib/apt/methods/https /usr/lib/apt/methods/https --process /tmp/runc-process3457108610 --detach --pid-file /run/containerd/io.containerd.runtime.v2.task/moby/08efd86bb6084c004dc18f20bcd8c26643dfdf24644c36ddda252b3e7c321a7e/8a4903d29032e6e0f7d9a7cbb56dbe15c4baa5d00ebc328013e361301a532f6e.pid 08efd86bb6084c004dc18f20bcd8c26643dfdf24644c36ddda252b3e7c321a7e sql/17/bin/psql /run/containerd/python 19e9ee20d3bb68af927c02fa91d4da420be/aabe5d2470e792854e99a86113d2f5b02d2a2cd9489b8677�� 732/log.json sql/17/bin/psql 732/�� y filter docker-buildx d73/config.json -i gko9qeip1ax docker-buildx (dns block)

If you need me to access, download, or install something from one of these locations, you can either:

• Configure Actions setup steps to set up my environment, which run before the firewall is enabled
• Add the appropriate URLs or hosts to the custom allowlist in this repository's Copilot coding agent settings (admins only)

Original prompt

Diagnostic + fix task for BGSTM#311

The E2E test environment is starting up with only requirements seeded — the 5 seeded test_cases (TC-001..TC-005) and their 3 requirement_test_case_links are missing from the database by the time the Traceability Matrix tests run. This causes frontend/tests/e2e/traceability.spec.ts:37 (seeded test case "TC-004" is visible in the matrix) to fail deterministically.

Full context, evidence, and three hypotheses are in BGSTM#311. Read that issue first — it has the failing-run screenshot evidence, the comparison table of observed vs. expected matrix metrics, and the diagnostic steps for each hypothesis.

Your job — diagnose, then fix

Step 1: Diagnose (don't skip)

Work through the three hypotheses in order and document what you find:

1. Hypothesis 1 — Schema mismatch in frontend/tests/e2e/fixtures/seed.sql.

   • Run seed.sql against a freshly migrated test DB locally (use the same Postgres image/version the CI compose file uses).
   • Capture stdout AND stderr. Look for any INSERT INTO test_cases ... ERROR — column count mismatch, NOT NULL violation, missing column, type mismatch, FK violation, etc.
   • Compare the column list in the INSERT INTO test_cases (...) statement against the actual test_cases table schema (check the latest Alembic migration that defines/alters it).
   • Same comparison for requirement_test_case_links.

2. Hypothesis 2 — backend/app/db/sample_data.py runs after seed.sql and overwrites/diverges.

   • Trace the test-environment startup sequence in docker-compose.test.yml (or whatever the E2E compose file is) and the backend container's entrypoint/start command.
   • Determine whether sample_data.py is invoked during E2E startup, and in what order relative to seed.sql loading.
   • If it runs, compare its TC-001..TC-00X UUIDs to those in seed.sql — ON CONFLICT (id) DO NOTHING only protects against re-inserting matching IDs.

3. Hypothesis 3 — A spec or fixture truncates test_cases.

   • git grep -n -E "TRUNCATE|DELETE FROM test_cases|truncate.*test_cases" frontend/ backend/
   • Check playwright.config.ts global setup, any beforeAll/globalSetup hooks, and backend test fixtures.

Document findings in the PR description as a "Diagnostic findings" section. State which hypothesis was confirmed (or describe the actual root cause if it turns out to be something else).

Step 2: Fix

Apply the minimal fix that addresses the confirmed root cause. Most likely candidates based on the evidence:

• If Hyp 1: correct the seed.sql column list / values to match the live schema. Add external_id and any other required columns. Keep the existing TC-001..TC-005 IDs and titles intact so existing tests still find them.
• If Hyp 2: disable sample_data.py for the E2E profile, or align its data with seed.sql's expected IDs.
• If Hyp 3: remove the offending truncation, or move it to only run before the spec that needs an empty state (and have that spec re-seed afterwards).

Step 3: Add a fail-fast smoke check

Per #311's acceptance criteria, add a smoke check to the E2E setup that fails fast with a clear error if seed loading is incomplete. Suggested approach: a tiny pre-test script (or a globalSetup step in playwright.config.ts) that queries the API or DB after seed and asserts row counts:

- requirements >= 5
- test_cases >= 5 (and TC-001..TC-005 specifically present)
- requirement_test_case_links >= 3

If any assertion fails, throw with a clear message naming which seed entity is missing. This prevents future regressions from masquerading as flaky UI tests for hours of debugging.

Step 4: Verify

• Re-run the full Playwright E2E suite locally — traceability.spec.ts:37 (TC-004) must pass on first attempt without retries.
• Traceability Matrix page shows non-zero coverage and all 5 seeded TC-00X rows appear in either the Linked Test Cases column (for the linked ones) or the Orphan Test Cases section (for TC-003 and TC-005, which are unlinked per seed.sql).
• CI passes on this PR's first attempt.

Constraints

• Do NOT mark the failing test fixme/skip. The point is to fix the underlying data-layer bug.
• Do NOT broaden scope into refactoring the entire seed strategy (raw SQL → programmatic, etc.) — that's explicitly out of scope per [Bug] E2E seed.sql silently fails to insert test_cases/links — Traceability Matrix shows 0% coverage, TC-004 missing #311.
• Do NOT touch the artifact-upload code from feat: artifact upload endpoint with pluggable storage backend (BGSTM #298) #310 or any external_results router code — unrelated.
• Keep the diff tightly scoped: ideally just seed.sql, possibly a startup script, plus the new smoke check.
• Reference [Bug] E2E seed.sql silently fails to insert test_cases/links — Traceability Matrix shows 0% coverage, TC-004 missing #311 in the PR description and link the closed/incorrect prior fix ([Test] E2E CRUD specs pollute seeded TC-00X rows, cause flaky Traceability Matrix failures #308) in a "Why [Test] E2E CRUD specs pollute seeded TC-00X rows, cause flaky Traceability Matrix failures #308 didn't fix this" sentence so future debuggers don't get misdirected.

Files likely involved

• frontend/tests/e2e/fixtures/seed.sql — primary suspect
• frontend/tests/e2e/playwright.config.ts (or equivalent) — for the new s...

This pull request was created from Copilot chat.