Summary

Stabilizes the session composer dock E2E coverage for #529 by replacing brittle scroll gestures and invalid one-option question seeds with deterministic test setup.

Why

The remaining #529 session-composer-dock slice was not a reliable user-path test:

• The dock height test tried to leave bottom-follow mode with a mouse wheel gesture that can be swallowed by nested scroll handling and leave the timeline at the bottom.
• Two question-dock tests seeded a single option even though the current question schema rejects one-option questions.
• The overflow keyboard test pressed End without focusing the question options, so it did not exercise the dock's keyboard path deterministically.

This PR keeps the product code unchanged and makes the existing E2E assertions match stable user paths.

Related Issue

Refs #529.

Human Review Status

Pending. A human should make the final merge decision after reviewing the final diff and verification evidence.

Review Focus

• The new seedSessionTurns() helper only creates deterministic scrollable history for this spec.
• scrollTimelineAwayFromBottom() now explicitly marks an upward wheel intent before moving the timeline, so the test exercises user-scrolled dock resize behavior instead of relying on flaky pointer placement.
• The question seeds now satisfy the current two-option minimum without changing the behavior under test.

Risk Notes

Low risk. This is E2E-only test stabilization with no product code, schema, dependency, or generated-file changes.

How To Verify

Red repro before fix: focused session-composer-dock run had 2 failures / 1 pass; failures were the dock-height scroll gesture and overflow question End-key path.
Focused E2E: 5 passed (session-composer-dock dock-height, latest question dock, overflow question dock, blocked prompt focus, multiline question text).
App typecheck: passed (bun run typecheck in packages/app).
Diff check: passed (git diff --check).

Screenshots or Recordings

Not applicable. E2E-only test stabilization, no visible product UI change.

Checklist

• Human review status is stated above as pending, approved, or not required
• I linked the related issue, or stated why there is no issue
• This PR has type, primary area, and priority labels, or I requested maintainer labeling
• I described the review focus and any meaningful risks
• I listed the relevant verification steps and the key result for each
• I did not introduce unrelated refactors, dependencies, generated files, or file changes beyond the stated scope
• I manually checked visible UI or copy changes when needed, with screenshots or recordings
• I considered macOS and Windows impact for platform, packaging, updater, signing, paths, shell, or permissions changes
• I called out docs, release notes, dependencies, permissions, credentials, deletion behavior, generated content, or local file changes when relevant
• I reviewed the final diff for unrelated changes and suspicious dependency changes
• I am targeting dev, and my PR title and commit messages use Conventional Commits in English

Summary by CodeRabbit

• Tests
  • Improved end-to-end test utilities and coverage for scroll behavior and keyboard interactions in the composer dock to enhance regression detection.