test: add tests for user activation vs file input interception

[Dmitry Gozman (dgozman)]

No description provided.

[github-actions]

Test results for "tests 1"

3 flaky

⚠️ [playwright-test] › ui-mode-trace.spec.ts:388 › should reveal errors in the sourcetab `@playwright-x64-server2025-64-core-node22`
⚠️ [firefox-library] › library/browsercontext-cookies-third-party.spec.ts:257 › third party 'Partitioned;' cookies `@firefox-ubuntu-22.04-node22`
⚠️ [firefox-library] › library/browsercontext-cookies-third-party.spec.ts:470 › top level 'Partitioned;' cookie and same origin iframe `@firefox-ubuntu-22.04-node22`

21775 passed, 525 skipped

---

Merge workflow run.

[github-actions]

Test results for "MCP"

3 failed
❌ [chromium] › mcp/http.spec.ts:515 › http transport shared context: one idle timer across clients @mcp-macos-latest-chromium
❌ [firefox] › mcp/cli-core.spec.ts:43 › click button @mcp-windows-latest-firefox
❌ [firefox] › mcp/cli-core.spec.ts:57 › click link @mcp-windows-latest-firefox

9201 passed, 1495 skipped

---

Merge workflow run.

[github-actions]

Hi, I'm the Playwright bot and I took a first look at the CI failures.

🟡 One failure is uncertain; the other two are pre-existing flakes

The Windows Firefox CLI failures have identical failures on unrelated SHAs. The macOS Chromium idle-timer failure does not appear connected to this test-only PR, but I could not confirm it as a flake from prior runs.

Details

This PR only adds file-chooser activation tests in tests/page/page-filechooser.spec.ts; it does not change MCP code or product code used by these failures.

Pre-existing flake / infra

• [firefox] › mcp/cli-core.spec.ts:43 › click button — pre-existing Windows Firefox flake. It failed 3 of 88 runs in the available history and passed the other 85. The same clickSnapshot-is-undefined failure occurred on unrelated sha 417ab248220d (run) and on fix(fill): accept range values the browser re-serializes #43343 (run).
• [firefox] › mcp/cli-core.spec.ts:57 › click link — pre-existing Windows Firefox flake. It failed 8 of 88 runs and passed the other 80. The same click timeout occurred on unrelated PRs, including fix(test): let registered loaders load files with unknown extensions #43305 (run), feat(test-runner): experimental esbuild transpilation #43320 (run), and fix(trace-viewer): fit the browser frame to phone-sized snapshots #43321 (run).

Uncertain

• [chromium] › mcp/http.spec.ts:515 › http transport shared context: one idle timer across clients — the log contains an unexpected browser restart while checking the shared idle timer. This exact macOS Chromium test passed 87 of 88 runs; its only failure in the available history is on this PR. The diff does not reach this MCP path, but without the same test failing on an unrelated SHA or passing on a rerun of this SHA, there is not enough evidence to call it a flake.

_{Triaged by the Playwright bot - agent run}