chore: re enable reasoning_content in backend

[qnixsynapse]

Describe Your Changes

Re enable reasoning_content in the llamacpp provider extension for a single standard API for rendering reasoning content

Fixes Issues

• Closes #
• Closes #

Self Checklist

• Added relevant comments, esp in complex areas
• Updated docs (for bug fixes / features)
• Created issues for follow-up changes or refactoring needed

---

Important

Re-enable reasoning_content in llamacpp extension, update reasoning extraction logic, and add tests for reasoning processing.

• Behavior:
  • Re-enable reasoning_content in AIEngine.ts and index.ts by removing --reasoning-format none.
  • Update extractReasoningFromMessage() in reasoning.ts to prefer reasoning_content over reasoning.
• Tests:
  • Add reasoning.test.ts to test reasoning extraction and processing.
  • Cover scenarios for reasoning_content, reasoning, and mixed content.
• Misc:
  • Add getReasoning() helper function in reasoning.ts for extracting reasoning content.

^{This description was created by for 52a8fd0. You can customize this summary. It will automatically update as commits are pushed.}

[ellipsis-dev]

Important

Looks good to me! 👍

Reviewed everything up to e0e58d6 in 57 seconds. Click for details.

• Reviewed 12 lines of code in 1 files
• Skipped 0 files when reviewing.
• Skipped posting 1 draft comments. View those below.
• Modify your settings and rules to customize what types of comments Ellipsis leaves. And don't forget to react with 👍 or 👎 to teach Ellipsis.

1. extensions/llamacpp-extension/src/index.ts:1212

• Draft comment:
  Removed the '--reasoning-format none' flag to re-enable reasoning content. Ensure the backend now defaults to producing reasoning output as expected and update docs/tests if necessary.
• Reason this comment was not posted:
  Comment did not seem useful. Confidence is useful = 0% <= threshold 50% The comment is asking the PR author to ensure that the backend defaults to producing reasoning output and to update docs/tests if necessary. This falls under the rule of not asking the author to ensure behavior or make sure changes are tested. Therefore, this comment should be removed.

Workflow ID: wflow_OAFP2HHpaQY7tha9

^{You can customize by changing your verbosity settings, reacting with 👍 or 👎, replying to comments, or adding code review rules.}

[github-actions]

Barecheck - Code coverage report

Total: 37.81%

Your code coverage diff: 0.22% ▴

✅ All code changes are covered

[ellipsis-dev]

Important

Looks good to me! 👍

Reviewed 2782b65 in 1 minute and 20 seconds. Click for details.

• Reviewed 431 lines of code in 3 files
• Skipped 0 files when reviewing.
• Skipped posting 3 draft comments. View those below.
• Modify your settings and rules to customize what types of comments Ellipsis leaves. And don't forget to react with 👍 or 👎 to teach Ellipsis.

1. core/src/browser/extensions/engines/AIEngine.ts:10

• Draft comment:
  Clarify that 'reasoning_content' takes precedence over the legacy 'reasoning'. Adding a note here can improve developer understanding.
• Reason this comment was not posted:
  Comment did not seem useful. Confidence is useful = 0% <= threshold 50% The comment is suggesting adding a note for clarification, which is a request for purely informative content. This violates the rule against making purely informative comments.

2. web-app/src/utils/reasoning.ts:43

• Draft comment:
  Consider documenting the behavior when a chunk contains both reasoning and content tokens. The current logic always prioritizes reasoning if present, which may be intentional but could use explicit clarification.
• Reason this comment was not posted:
  Comment was not on a location in the diff, so it can't be submitted as a review comment.

3. web-app/src/utils/reasoning.ts:36

• Draft comment:
  For consistency, consider renaming the internal state variable 'isReasoningActive' to something like 'isReasoningInProgress' to match the getter method name.
• Reason this comment was not posted:
  Comment was not on a location in the diff, so it can't be submitted as a review comment.

Workflow ID: wflow_9hnq3DrnGLP5lmqx

^{You can customize by changing your verbosity settings, reacting with 👍 or 👎, replying to comments, or adding code review rules.}

[dinhlongviolin1]

LGTM @qnixsynapse @urmauur

[ellipsis-dev]

Caution

Changes requested ❌

Reviewed ff63e33 in 1 minute and 50 seconds. Click for details.

• Reviewed 34 lines of code in 1 files
• Skipped 0 files when reviewing.
• Skipped posting 2 draft comments. View those below.
• Modify your settings and rules to customize what types of comments Ellipsis leaves. And don't forget to react with 👍 or 👎 to teach Ellipsis.

1. web-app/src/utils/reasoning.ts:20

• Draft comment:
  Refactor: Using getReasoning in extractReasoningFromMessage cleans up the logic. This improves maintainability.
• Reason this comment was not posted:
  Confidence changes required: 0% <= threshold 50% None

2. web-app/src/utils/reasoning.ts:30

• Draft comment:
  Consistently applying getReasoning in extractReasoningFromChunk ensures uniform handling of reasoning content across different contexts.
• Reason this comment was not posted:
  Confidence changes required: 0% <= threshold 50% None

Workflow ID: wflow_7FJ6tFhJfQ8dHmC1

^{You can customize by changing your verbosity settings, reacting with 👍 or 👎, replying to comments, or adding code review rules.}

[ellipsis-dev]

Important

Looks good to me! 👍

Reviewed 52a8fd0 in 1 minute and 22 seconds. Click for details.

• Reviewed 13 lines of code in 1 files
• Skipped 0 files when reviewing.
• Skipped posting 1 draft comments. View those below.
• Modify your settings and rules to customize what types of comments Ellipsis leaves. And don't forget to react with 👍 or 👎 to teach Ellipsis.

1. web-app/src/utils/reasoning.ts:10

• Draft comment:
  Using the nullish coalescing operator (??) here is intentional to prefer a provided empty string over falling back to 'reasoning'. Confirm that empty strings in 'reasoning_content' should be preserved rather than triggering the fallback.
• Reason this comment was not posted:
  Comment looked like it was already resolved.

Workflow ID: wflow_1AKB9p2GzgWg9qJ9

^{You can customize by changing your verbosity settings, reacting with 👍 or 👎, replying to comments, or adding code review rules.}