fix(openai-chat): keep captured usage on an invalid negotiated completion by xubenqiang · Pull Request #7849 · lidge-jun/opencodex

xubenqiang · 2026-10-27T19:39:01Z

Every terminal path of the OpenAI-compatible Chat streaming generator attaches pendingUsage to its error event — invalidChoicesEvent(pendingUsage), invalidToolCallsEvent(..., pendingUsage), upstreamErrorEvent(..., pendingUsage) — except the invalid negotiated tool completion path (block-stop-v1, #7561), which built its event without it. When upstream reported usage before a repeated completion marker or a late argument fragment, those tokens were missing from the request's error event.

Change

src/adapters/openai-chat.ts

The completion === "invalid" branch now passes usage: pendingUsage (when captured) like its sibling terminal errors.
No other behaviour changes: the event type, message and status are the same.

Test

tests/adapters/openai/openai-chat-parallel-stream.test.ts

New an invalid completion sequence still reports the usage already captured: streams a call, its completion marker, a usage-only chunk, then a late argument fragment that trips observeCompletedToolBlock, and asserts the final error event is still the negotiated-completion error and carries { inputTokens: 11, outputTokens: 7 }.

Verified: bun test tests/adapters/openai/openai-chat-parallel-stream.test.ts -> 52 pass / 0 fail; with only openai-chat.ts reverted, 51 pass / 1 fail (the new test). bun x tsc --noEmit exits 0.

Summary by CodeRabbit

Bug Fixes
- Error reports for invalid tool-completion sequences now include captured input and output token usage when available, making usage information accessible even when a stream ends with an error. This improves visibility into token consumption in these cases without changing the behavior of valid tool-completion sequences.

Review readiness checklist

This PR stays in draft until every box below is ticked. Tick all four boxes once the requirements are met:

Required local validation passed; commands, results, and any full-suite exception are documented.
I pushed my PR to a recent dev commit (at most 10 behind; a maintainer may still ask for the exact tip before merge).
I resolved all correct Codex and CodeRabbit findings.
My PR is ready for review.

…tion Every other terminal path of the streaming generator attaches `pendingUsage` to its error event (`invalidChoicesEvent`, `invalidToolCallsEvent`, `upstreamErrorEvent`), but the invalid negotiated tool completion path threw it away. When upstream reported usage before a repeated completion marker or a late argument fragment, those tokens were missing from the request's error event even though every other error carries them. Pass `pendingUsage` through like the sibling terminal errors.

Copilot

Copilot was unable to review this pull request because the user who requested the review has insufficient remaining quota.

coderabbitai · 2026-10-27T19:39:30Z

Caution

Review failed

The pull request is closed.

No actionable comments were generated in the recent review.

Recent review info

Run configuration

Source mode: Auto
Repository: lidge-jun/opencodex
Branch: dev
File: .coderabbit.yaml
Commit: 0c32c81
Layers: Repository YAML
Review profile: ASSERTIVE
Plan: Advanced
Run ID: b9babe63-abea-4b69-a5a7-aeefbfeda037

Commits

Reviewing files that changed from the base of the PR and between 0c32c81 and 9d91a2e.

Files selected for processing: 2

src/adapters/openai-chat.ts
tests/adapters/openai/openai-chat-parallel-stream.test.ts

Included review availability: This review used your included allowance. Your plan provides up to 10 included reviews per hour; 9 remain after this review.

Walkthrough

When a negotiated tool-completion sequence is invalid, the stream error now includes captured usage when available. A regression test checks that the error retains the input and output token counts.

Changes: Usage in stream errors

Layer / File(s)	Summary
Include pending usage in invalid completion errors `src/adapters/openai-chat.ts`, `tests/adapters/openai/openai-chat-parallel-stream.test.ts`	The invalid negotiated tool-completion error now includes pending usage. The regression test checks that the error contains `inputTokens: 11` and `outputTokens: 7`.

Priority: Low

Estimated code review effort: 2 (Simple) | ~8 minutes

Change: Bug fix · Severity of issue fixed: Low

Suggested labels: bug

Suggested reviewers: lidge-jun

Merge Risk: Minimal · previous review 9d91a

The PR is mergeable with no actionable risks identified.

Pre-merge checks: 4 · 1

Failed checks: 1 warning

Check name	Status	Explanation	Resolution
Docstring Coverage		Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Documentation-only edits and deletions within survi…	Write docstrings for the functions missing them to satisfy the coverage threshold.

Passed checks: 4 passed

Check name	Status	Explanation
Description Check		Check skipped - CodeRabbit’s high-level summary is enabled.
Title check		The title clearly and concisely describes the main change: preserving captured usage for invalid negotiated completions in the OpenAI Chat adapter.
Linked Issues check		Issue `#7841` requires the invalid negotiated tool-completion path to preserve upstream usage and requires a regression test. In `src/adapters/openai-chat.ts`, the `completion === "invalid"` path now su…
Out of Scope Changes check		The reported pull request changes are limited to the required adapter fix in `src/adapters/openai-chat.ts` and its regression test in `tests/adapters/openai/openai-chat-parallel-stream.test.ts`. The t…

Full details: Linked Issues check

Explanation

Issue #7841 requires the invalid negotiated tool-completion path to preserve upstream usage and requires a regression test. In src/adapters/openai-chat.ts, the completion === "invalid" path now supplies usage: pendingUsage to the existing negotiated-completion error event. The event type, message, and status remain unchanged. In tests/adapters/openai/openai-chat-parallel-stream.test.ts, the new test sends completion, usage, and a late argument fragment, then asserts { inputTokens: 11, outputTokens: 7 } on the error event. The supplied test result reports 52 passes and 0 failures.

Full details: Out of Scope Changes check

Explanation

The reported pull request changes are limited to the required adapter fix in src/adapters/openai-chat.ts and its regression test in tests/adapters/openai/openai-chat-parallel-stream.test.ts. The test directly exercises the repeated-completion or late-fragment failure mode from issue #7841. No unrelated behavior, public declaration, or feature change is identified.

Full details: Docstring Coverage

Explanation

Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Documentation-only edits and deletions within surviving functions are included; unchanged diff context and fully deleted functions are excluded. Analyzed 1 functions across 2 files.

Resolution

Write docstrings for the functions missing them to satisfy the coverage threshold.

Finishing Touches

🧪 Generate unit tests (beta)

Create a new PR

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.

❤️ Share

_{Comment @coderabbitai help to get the list of available commands.}

github-actions · 2026-10-27T21:07:20Z

✅ Deterministic PR hygiene checks passed.

github-actions · 2026-10-28T02:02:18Z

⏳ DRAFT

review readiness checklist open (0/4 boxes ticked).

What to do

Tick all four boxes in the PR description once you're done (currently 0/4).

Review readiness checklist

⬜ Required local validation passed; commands, results, and any full-suite exception are documented.
⬜ I pushed my PR to a recent dev commit (at most 10 behind; a maintainer may still ask for the exact tip before merge).
⬜ I resolved all correct Codex and CodeRabbit findings.
⬜ My PR is ready for review.

0/4 boxes ticked.

Automatic draft conversion failed. Please convert this pull request to a draft manually until every box above is ticked.

Hygiene

✅ Deterministic PR hygiene checks passed.

lidge-jun · 2026-10-28T02:36:55Z

Maintainer integration into dev (MAINTAINERS.md, "Review and merge policy"). This is maintainer integration, not an approving review.

Exact head 4945dd4914ffe14272650a19d15c3cc05115ebc9; no conflict with current dev.
Maintainer change on top of the author's fix: 4945dd491 folds the new usage spread into the existing single-line return. src/adapters/openai-chat.ts sits at its file-size cap, and the first head grew it by six lines, which failed test changed (file-size ratchet). Behavior is unchanged.
Exact-head PR CI: Cross-platform CI 40620184440 (gates, test changed, ci), Catalog async contracts 40620184382, enforce-target: all success.
Local (Bun 1.4.4): openai-chat-parallel-stream + file-size ratchet 61 pass / 0 fail; the new regression fails without the source change (per the description, 51 pass / 1 fail).

Fixes #7841. Thank you @xubenqiang.

chatgpt-codex-connector · 2026-10-28T02:37:26Z

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	🔄 Running since 2026-10-28T02:37:21.555936Z	`4945dd4`	Draft marked ready

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

Copilot AI balanced review requested due to automatic review settings October 27, 2026 19:39

xubenqiang requested review from Ingwannu and lidge-jun as code owners October 27, 2026 19:39

Copilot AI reviewed Oct 27, 2026

github-actions Bot added the bug Something isn't working label Oct 27, 2026

github-actions Bot marked this pull request as draft October 28, 2026 02:02

fix(openai-chat): keep the usage branch within the file-size cap

4945dd4

lidge-jun marked this pull request as ready for review October 28, 2026 02:37

lidge-jun merged commit 1997c69 into lidge-jun:dev Oct 28, 2026
27 checks passed

lidge-jun mentioned this pull request Oct 28, 2026

Chat stream error for an invalid negotiated tool completion omits captured usage #7841

Closed

2 tasks done

coderabbitai Bot mentioned this pull request Oct 28, 2026

[Compatibility] OpenDesign DeepSeek v4.1 streaming tool arguments fail with 502 via combo; opencode-go route works #7920

Open

3 tasks done

Navigation Menu

Repository navigation

fix(openai-chat): keep captured usage on an invalid negotiated completion - #7849

fix(openai-chat): keep captured usage on an invalid negotiated completion#7849
lidge-jun merged 2 commits into
lidge-jun:devfrom
xubenqiang:fix/negotiated-completion-usage

xubenqiang commented Oct 27, 2026 •

edited by github-actions Bot

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

coderabbitai Bot commented Oct 27, 2026 •

edited

Loading

Review failed

Uh oh!

github-actions Bot commented Oct 27, 2026

Uh oh!

github-actions Bot commented Oct 28, 2026 •

edited

Loading

Uh oh!

lidge-jun commented Oct 28, 2026

Uh oh!

Uh oh!

chatgpt-codex-connector Bot commented Oct 28, 2026

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

xubenqiang commented Oct 27, 2026 • edited by github-actions Bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Change

Test

Summary by CodeRabbit

Review readiness checklist

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Uh oh!

coderabbitai Bot commented Oct 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Review failed

Uh oh!

github-actions Bot commented Oct 27, 2026

Uh oh!

github-actions Bot commented Oct 28, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

⏳ DRAFT

What to do

Review readiness checklist

Hygiene

Uh oh!

lidge-jun commented Oct 28, 2026

Uh oh!

Uh oh!

chatgpt-codex-connector Bot commented Oct 28, 2026

Codex Review Summary

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

xubenqiang commented Oct 27, 2026 •

edited by github-actions Bot

Loading

coderabbitai Bot commented Oct 27, 2026 •

edited

Loading

github-actions Bot commented Oct 28, 2026 •

edited

Loading