Streaming path: preserve provider token usage - #50
Merged
Conversation
executeTurn's streaming branch discarded the usage on the final `.finish` chunk (`case .finish(let reason, _)`) and synthesized the response with `usage: nil`, forcing cost/context onto a local estimate. Capture the usage and thread it onto the returned response (text and tool-call paths). Adds ReplayRun.runStreaming to exercise the streaming path, and a regression test asserting the emitted response carries the provider's promptTokens/completionTokens. Co-Authored-By: Claude Opus 4.8 (1M context) <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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What
The streaming turn (
executeTurnwithonText != nil) discarded the provider's token usage on the final.finishchunk (case .finish(let reason, _)) and built the response withusage: nil. That forced consumers (cost tracking, context indicators) onto a local character-based estimate instead of the model's real reported tokens.Fix
usagefrom the.finish(reason:usage:)chunk intostreamedUsage.complete) already carried usage — unchanged.Tests
ReplayRun.runStreaminghelper to exercise the streaming path.StreamingUsageTests.streamingResponseCarriesProviderUsage— asserts the emitted.llmCallCompletedresponse carries the provider'spromptTokens/completionTokens. Verified RED without the fix (usage wasnil), GREEN with it.🤖 Generated with Claude Code