Fix/thinking UI app

[SOUMITRO-SAHA]

Issue for this PR

Closes #21548

Type of change

• Bug fix
• New feature
• Refactor / code improvement
• Documentation

What does this PR do?

Implements a collapsible "Thinking" container for model reasoning outputs in the UI.

How did you verify your code works?

• Ran bun typecheck in packages/opencode - passed
• Ran linting - passed
• Manual testing: toggled /thinking command and verified reasoning containers open/close correctly
• Verified setting persistence across sessions

Screenshots / recordings

Checklist

• I have tested my changes locally
• I have not included unrelated changes in this PR

[github-actions]

Hey! Your PR title Fix/thinking UI app doesn't follow conventional commit format.

Please update it to start with one of:

• feat: or feat(scope): new feature
• fix: or fix(scope): bug fix
• docs: or docs(scope): documentation changes
• chore: or chore(scope): maintenance tasks
• refactor: or refactor(scope): code refactoring
• test: or test(scope): adding or updating tests

Where scope is the package name (e.g., app, desktop, opencode).

See CONTRIBUTING.md for details.

[github-actions]

The following comment was made by an LLM, it may be inaccurate:

I found potential related PRs:

1. PR feat(desktop): Make Reasoning Collapsable #15259: "feat(desktop): Make Reasoning Collapsable"

   • feat(desktop): Make Reasoning Collapsable #15259
   • This appears to be directly related - it implements collapsible reasoning blocks, which is exactly what PR Fix/thinking UI app #21549 is doing for the thinking UI.

2. PR feat(ui): Add thinking/reasoning block visibility toggle #9139: "feat(ui): Add thinking/reasoning block visibility toggle"

   • feat(ui): Add thinking/reasoning block visibility toggle #9139
   • Related feature for toggling thinking/reasoning block visibility in the UI.

3. PR fix(provider): preserve assistant message content when reasoning blocks present #21370: "fix(provider): preserve assistant message content when reasoning blocks present"

   • fix(provider): preserve assistant message content when reasoning blocks present #21370
   • Related to handling reasoning blocks, though this is more provider-focused.

The most potentially overlapping is PR #15259 which appears to address the same feature (making reasoning/thinking blocks collapsible). You may want to check if that PR was closed/merged and whether this is a reopened or complementary fix.

[SOUMITRO-SAHA]

I'm not entirely confident that this PR will be merged. Over time, additional changes may introduce merge conflicts, which could further reduce its chances of being accepted.

As a result, I've decided to maintain my own fork of opencode. This message is for anyone interested in trying out these features—you can visit my fork and refer to the "Notes" section in the README.md for detailed guidance.

The setup process is straightforward and easy to follow.

Fork: https://github.com/SOUMITRO-SAHA/opencode

use opencode on a daily basis and will continue to resolve all issues that I regularly face. Feel free to create new issues in my fork—I'll make sure they're resolved as soon as possible.
You can also bring issues from the original opencode repository to my fork's issue section—I'll surely look into those as well.

Issue Tracker: https://github.com/SOUMITRO-SAHA/opencode/issues