Repository navigation
Conversation
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info
📝 Walkthrough
Priority: ➖ Normal Merge Risk: ⚪ Minimal · up to Completed responses remain visible, and early task checks continue polling until the result is available. No merge-blocking issue was established. Pre-merge checks |
  |
| Check name | Status | Explanation | Resolution |
|---|---|---|---|
| Description check |  |
The description only records AI-generation metadata. It does not describe the chat streaming or polling changes. | Add a short summary of the changes, such as progressive streamed-text reveals, immediate completion checks after push events, and protection against overlapping polling requests. |
✅ Passed checks (4 passed)
-
Check name Status Explanation Docstring Coverage Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 1 functions across 2 files. (3 skipped: 3 … Linked Issues check Check skipped because no linked issues were found for this pull request. Out of Scope Changes check Check skipped because no linked issues were found for this pull request. Title check The title clearly summarizes the main changes: faster streamed answer display and completion on push events.
- Fix all pre-merge checks with AI
- Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts
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.
Comment @coderabbitai help to get the list of available commands.
ifedan-ed
force-pushed
the
enh/noid/faster-chat-streaming
branch
from
October 10, 2026 22:56
7b7b783 to
93f8a5e
Compare
added 2 commits
October 11, 2026 01:00
The streaming chat message typed the received text one character every 5 ms and re-rendered the markdown after each character. That caps the display at 200 characters per second, slower than common models stream (about 140 tokens/s, roughly 550 characters/s, for a flash model), so the visible text fell behind the received text and jumped ahead with each new chunk. The first chunk was not displayed at all: the watcher was not immediate, and the streaming message is mounted with the first chunk already set, so it showed "Reasoning…" until the second chunk. Reveal each new chunk over at most 200 ms in ~16 ms steps. That is shorter than the 250 ms interval at which integration_openai reports intermediate output, so the display catches up before the next chunk. Show the content that was received before the component was mounted. Signed-off-by: Daniel <danitex@danvics.com>
The chat checked the generation task only every 2 seconds, so the final message and the re-enabled input came 0 to 2 s (1 s on average) after the task had finished, even when notify_push had already announced it. assistant.js already turns the taskprocessing:task_update push into assistant:task:updated and assistant:task:status:updated events. Listen to them while polling and check right away when the task succeeded, failed or was cancelled. Polling every 2 s stays as the fallback without notify_push. Only one check request runs at a time. Signed-off-by: Daniel <danitex@danvics.com>
ifedan-ed
force-pushed
the
enh/noid/faster-chat-streaming
branch
from
October 10, 2026 23:00
93f8a5e to
26d3dc0
Compare
ifedan-ed
changed the title
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
🤖 AI (if applicable)