fix(app): make model picker height adaptive to model count

[Astro-Han]

Summary

Switch the model selector popover from a fixed h-[400px] to max-h-[400px] so it hugs its natural content height and only scrolls when models exceed the 400px cap. Add an E2E spec that locks the geometry invariant.

Why

When a user only has a handful of visible models (e.g. the OpenCode Zen free tier with four entries), the popover was still forced to 400px tall, leaving a ~190px empty band between the last model row and the Thinking row at the bottom. The popover should adapt to model count.

Root cause: packages/app/src/components/prompt-input/model-picker.tsx:218 set a fixed height on the Kobalte.Content. The inner ModelList uses flex-1 min-h-0, so it expanded to fill the fixed 400px and produced a visible empty band above the Thinking row.

Related Issue

Closes #719.

Human Review Status

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

Review Focus

• The single CSS class change on Kobalte.Content. The flex-1 min-h-0 chain on ModelList -> list-scroll still works under max-h because the chain stays unbroken when the cap kicks in; when content fits, the column shrinks to content height instead.
• The new E2E spec: it asserts the fixture renders a short list that fits without scrolling (precondition), then asserts there is no empty band between the last [data-slot="list-item"] and [data-action="prompt-model-thinking-trigger"], and finally that popover.height <= 400. If a future fixture grows large enough to overflow the cap, the precondition fails loudly rather than silently skipping the regression check.
• The spec is tagged @smoke so the regression is enforced by the PR e2e-artifacts gate. The smoke inventory in packages/opencode/test/config/e2e-smoke-tagging.test.ts is updated in lockstep (alphabetical order).

Risk Notes

None. The fix is a single CSS class swap with no behavioral side effects. The new spec runs in the existing Playwright project, no infra changes.

How To Verify

Negative check (revert fix, run spec): 1 failed — gap below last model = 193.25px
Positive check (apply fix, run spec):  1 passed — popover hugs content
bun --cwd packages/app run test:e2e:local e2e/model-picker-height.spec.ts

Smoke inventory contract: bun --cwd packages/opencode test test/config/e2e-smoke-tagging.test.ts -> 2 pass.

Screenshots or Recordings

E2E captures a screenshot artifact at e2e/test-results/.../model-picker.png on every run. With the fix applied the popover ends right under the four visible models and the Thinking row, with no empty band. Local screenshot reviewed and confirmed visually before opening this PR.

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 exactly one type label (bug, enhancement, task, or documentation), at least one primary routing label (app, ui, platform, harness, or ci), and exactly one priority label (P0 to P3), 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

[coderabbitai]

Warning

Rate limit exceeded

@Astro-Han has exceeded the limit for the number of commits that can be reviewed per hour. Please wait 54 minutes and 16 seconds before requesting another review.

You’ve run out of usage credits. Purchase more in the billing tab.

⌛ How to resolve this issue?

After the wait time has elapsed, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

We recommend that you space out your commits to avoid hitting the rate limit.

🚦 How do rate limits work?

CodeRabbit enforces hourly rate limits for each developer per organization.

Our paid plans have higher rate limits than the trial, open-source and free plans. In all cases, we re-allow further reviews after a brief timeout.

Please see our FAQ for further information.

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro Plus

Run ID: 6e45b38c-571c-4c69-bfa0-5cc6a79576f5

📥 Commits

Reviewing files that changed from the base of the PR and between b57d673 and 2c1b53a.

📒 Files selected for processing (3)

• packages/app/e2e/model-picker-height.spec.ts
• packages/app/src/components/prompt-input/model-picker.tsx
• packages/opencode/test/config/e2e-smoke-tagging.test.ts

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch claude/fix-model-picker-height

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[github-actions]

Suggested priority: P2 (includes user-path files (packages/app/src/components/prompt-input/model-picker.tsx)).

P1/P0 are reserved for maintainer confirmation. Please relabel manually if this is a release blocker, security issue, data-loss risk, or updater/runtime failure.

[gemini-code-assist]

Code Review

This pull request updates the model picker's styling to use a maximum height instead of a fixed height, ensuring the container fits its content without leaving empty space. A new E2E test was added to validate this behavior. The review feedback suggests improving the robustness of the E2E test by using a more specific selector to avoid ambiguity and potential fragility.

[github-actions]

Perf delta summary

Comparator: pass

Profile / Scenario	interaction median	interaction worst	long task max	tbt	frame gap p95	frame gap max	jank count	cls	status
default / homepage-cold	24 -> 24 (0)	56 -> 48 (-8)	75 -> 66 (-9)	25 -> 16 (-9)	16.8 -> 16.8 (0)	166.6 -> 116.7 (-49.9)	4 -> 3 (-1)	0 -> 0 (0)	pass
default / long-session-input-lag	48 -> 48 (0)	48 -> 64 (+16)	0 -> 0 (0)	0 -> 0 (0)	16.7 -> 16.7 (0)	16.8 -> 16.8 (0)	0 -> 0 (0)	0 -> 0 (0)	pass
default / session-streaming-long	48 -> 40 (-8)	64 -> 64 (0)	0 -> 0 (0)	0 -> 0 (0)	16.8 -> 16.8 (0)	33.4 -> 16.8 (-16.6)	0 -> 0 (0)	0 -> 0 (0)	pass
default / tool-call-expand	16 -> 24 (+8)	40 -> 40 (0)	0 -> 0 (0)	0 -> 0 (0)	16.7 -> 16.8 (+0.1)	16.7 -> 16.8 (+0.1)	0 -> 0 (0)	0 -> 0 (0)	pass
default / tool-default-open-heavy-bash	24 -> 32 (+8)	24 -> 40 (+16)	64 -> 61 (-3)	14 -> 11 (-3)	50 -> 50 (0)	116.7 -> 100 (-16.7)	3 -> 2 (-1)	0 -> 0 (0)	pass
default / terminal-side-panel-open	48 -> 48 (0)	64 -> 56 (-8)	0 -> 0 (0)	0 -> 0 (0)	33.3 -> 33.4 (+0.1)	33.3 -> 33.4 (+0.1)	0 -> 0 (0)	0 -> 0 (0)	pass
default / session-scroll-reading	32 -> 24 (-8)	32 -> 32 (0)	0 -> 0 (0)	0 -> 0 (0)	16.7 -> 16.7 (0)	16.7 -> 16.7 (0)	0 -> 0 (0)	0.505 -> 0.505 (0)	warn: cls

[Astro-Han]

Thanks for the audit. Addressed the two P3 items:

• P3-1 (silent skip of gap assertion): Fixed in cc10500. The if (!listOverflowing) guard is gone — the spec now asserts scrollHeight <= clientHeight + 1 as a precondition, so if a future fixture grows large enough to overflow the 400px cap the spec fails loudly instead of skipping the regression check.
• P3-2 (not in @smoke): Decided against adding @smoke. This is a low-frequency visual regression guard, not a critical-path smoke check, and widening the PR smoke surface costs CI time on every PR. Full E2E sweeps will catch regressions. PR body now states this explicitly.

unit-opencode and all other CI jobs are green at the latest commit.

[Astro-Han]

Reversing my earlier call on P3-2. After a second pass through the three-question check:

• "Could this be even less?" — I was wrong to treat "skip the smoke gate" as the cheaper path. The actual cost is ~4 seconds per PR plus a one-time inventory entry; the cost of not gating is that a regression of this real, user-reported visible bug can slip past PR check and reach a release cut before nightly E2E catches it.
• "Is what remains good enough?" — No. The good-enough bar for a regression of a real user-reported visible bug is "don't let it reach users again", not "nightly will eventually find it".
• "Is it reassuring?" — Future maintainers touching model-picker.tsx benefit from PR-time feedback, not next-day feedback.

Tagged @smoke and synced the inventory in alphabetical order in ae1007f5. PR body updated.