fix(ci): stop the E2E shard from being cancelled mid-run (timeout headroom) (#3387)

The heaviest E2E shard (5/6 — responsive viewport matrix + studio/smoke) overran
the job's 20m timeout-minutes because each shard re-runs `npm run build` (~5m)
before Playwright, then runs ~24 serial tests with retries:2. The job was killed
(CANCELLED mid-run, 'Terminate orphan process') instead of any test failing.

- Bump test-e2e timeout-minutes 20 -> 35 (cumulative build+tests headroom).
- Lower the Playwright per-test timeout 600s -> 180s so a genuine hang fails fast
  and visibly (a clear per-test timeout) instead of silently eating the job budget.

.github/workflows/ci.yml

@@ -500,7 +500,11 @@ jobs:
   test-e2e:
     name: E2E Tests (${{ matrix.shard }}/6)
     runs-on: ubuntu-latest
-    timeout-minutes: 20
+    # 20m was too tight for the heaviest shard: each shard re-runs `npm run build`
+    # (~5m) before Playwright, so a slow shard's tests + build overran 20m and the
+    # job was cancelled mid-run. 35m gives headroom; the per-test cap in
+    # playwright.config.ts bounds any genuine hang to a fast, visible failure.
+    timeout-minutes: 35
     needs: build
     strategy:
       fail-fast: false

playwright.config.ts

@@ -25,7 +25,11 @@ export default defineConfig({
     "**/skills-marketplace.spec.ts",
   ],
   fullyParallel: false,
-  timeout: 600_000,
+  // Per-test cap. 600s was high enough that one hung test (× retries) could
+  // exhaust the e2e job's wall-clock budget, so the GitHub job hit its
+  // timeout-minutes and was CANCELLED mid-run instead of the test failing fast.
+  // 180s is generous for a UI flow yet bounds a hang to a clear per-test failure.
+  timeout: 180_000,
   forbidOnly: !!process.env.CI,
   retries: process.env.CI ? 2 : 0,
   workers: 1,