fix(context_engine): preserve reasoning fields in history projection#101
Open
0xKT wants to merge 1 commit into
Open
fix(context_engine): preserve reasoning fields in history projection#1010xKT wants to merge 1 commit into
0xKT wants to merge 1 commit into
Conversation
DeepSeek thinking mode rejects a follow-up turn whose replayed assistant message carried a tool call: "reasoning_content must be passed back to the API". The two history-projection allowlists dropped reasoning_content and thinking_blocks, so the rebuilt history lost them. Add both keys to _history_from_messages (Curator) and _ALLOWED_KEYS (HistoryTrimmer); the provider gate still strips thinking_blocks for non-Anthropic targets. Also count both fields in estimate_prompt_tokens and estimate_message_tokens so the preserved payload does not overflow the trim budget on long multi-turn thinking chains. Co-authored-by: Claude (claude-opus-4-8) <noreply@anthropic.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Change description
DeepSeek thinking mode rejects a follow-up turn whose replayed assistant message carried a tool call, with "reasoning_content must be passed back to the API". The two history-projection allowlists dropped
reasoning_content/thinking_blocks, so the rebuilt assistant history lost them.Both keys are now preserved in
_history_from_messages(Curator) and_ALLOWED_KEYS(HistoryTrimmer); the provider gate still stripsthinking_blocksfor non-Anthropic targets. The token estimators (estimate_prompt_tokens/estimate_message_tokens) now count both fields as well, so the preserved payload does not overflow the trim budget on long multi-turn thinking chains.Type of change
Related issues
Fixes #89
Checklists
Development
Security
Code review