updated read me to include vault instructions - #3
Merged
Merged
Conversation
damianloch
requested changes
Jan 23, 2026
|
|
||
| # Restart Aurora to load the Vault token | ||
| make down | ||
| make dev |
Contributor
There was a problem hiding this comment.
make prod-local this was previous error
4 tasks
6 tasks
damianloch
added a commit
that referenced
this pull request
May 11, 2026
- Extract _extract_reasoning_from_delta() helper to reduce cognitive complexity (SonarQube S3776: was 33, threshold 15) - Track has_reasoning_details separately so reasoning-only chunks with empty text entries still get content cleared (CodeRabbit #1) - Remove `if not reasoning` gate in workflow.py streaming — upstream _ReasoningChatOpenAI already clears content on reasoning-only chunks, so mixed chunks with both reasoning and visible content are handled correctly (CodeRabbit #3)
damianloch
added a commit
that referenced
this pull request
May 12, 2026
* fix: prevent reasoning tokens from streaming as normal output Google models via OpenRouter (and direct Gemini) were leaking reasoning/ thinking tokens into the frontend chat as if they were regular output. Root causes: - _ReasoningChatOpenAI only checked delta.reasoning but Google models via OpenRouter use delta.reasoning_details (array format) - workflow.py used include_thinking=True unconditionally, extracting thinking blocks from Gemini's list-format content for all chats - No guard against reasoning-only chunks duplicating into content Fixes: - Expand _ReasoningChatOpenAI to parse reasoning_details array and clear content on reasoning-only chunks - Add reasoning.exclude=true for Google models in user-facing chats (model still reasons internally, just doesn't return the tokens) - Use include_thinking=False in streaming path; only forward reasoning to background RCA chats for incident thought tracking * address PR review feedback - Extract _extract_reasoning_from_delta() helper to reduce cognitive complexity (SonarQube S3776: was 33, threshold 15) - Track has_reasoning_details separately so reasoning-only chunks with empty text entries still get content cleared (CodeRabbit #1) - Remove `if not reasoning` gate in workflow.py streaming — upstream _ReasoningChatOpenAI already clears content on reasoning-only chunks, so mixed chunks with both reasoning and visible content are handled correctly (CodeRabbit #3) * fix: resolve SonarCloud code smells in touched files - Replace logging.error with logging.exception in except blocks (S8572) - Replace chained startswith calls with tuple arg (S8513) * fix: prevent thinking blocks from leaking into persisted messages _consolidate_message_chunks and _convert_to_ui_messages were still using include_thinking=True, causing thinking blocks from direct Google models to survive into chat_sessions.messages and appear on page reload. * Revert "fix: prevent thinking blocks from leaking into persisted messages" This reverts commit 84b827d. * fix: prevent thinking blocks from leaking into persisted messages - _consolidate_message_chunks and _convert_to_ui_messages now use include_thinking=False so thinking blocks from direct Google models don't survive into chat_sessions.messages on page reload. - Added clarifying comment on OpenRouter exclude vs direct SDK filtering. * fix: preserve thinking tokens for background RCA while filtering in foreground Use is_background to control include_thinking in the streaming path so direct-mode OpenAI/Anthropic reasoning flows into incident thoughts for RCA chats but stays hidden for user-facing conversations.
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
2 participants
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.
small fix relating to read me instructions