[HAYDEN-OAI]

Summary

Make examples/responses/stream.ts and streaming-tools.ts report unsuccessful final responses instead of exiting successfully. Draining a stream does not imply that its response completed: finalResponse() deliberately resolves the latest snapshot for failed or incomplete responses and after a clean EOF with an in-progress snapshot.

After draining, check the final response status, emit a status-only diagnostic, and set process.exitCode = 1 unless it is completed. Keep existing event/final-response stdout, request parameters, tool schema, and normal rejected-request behavior unchanged. The SDK's finalResponse() contract is not changed, and no retry/resume behavior is added.

Reuse the existing Responses example integration harness for eight additional real-entrypoint cases: completed, failed, incomplete, and clean EOF after response.created for each example. Fixtures retain runtime/proxy/trust settings, use a synthetic API key and loopback endpoint, and verify request count, model, and the existing strict query tool.

Validation

• Exact final tests against unchanged examples: six expected exit-status failures; both completed controls and all 26 existing background/guide cases pass. Request and authentication assertions pass before each failing exit-status assertion.
• All 34 focused integration tests pass on Node 22.22.3, 24.19.0, and 26.7.0. For the local Node 22.22.3 run, Node warnings were redirected to a temporary log because the environment's proxy preload emits an experimental warning; runtime/proxy/trust settings and test assertions were unchanged. The log contains only that known warning.
• All eight actual example-entrypoint checks pass on the exact Node 22.0.0 floor, with natural exits and one synthetic-authenticated loopback request each.
• Full handwritten suite: 7,707 tests across 203 files pass on Node 24.19.0.
• Canonical formatting, lint, full repository TypeScript check, and git diff --check pass.
• Adversarial self-review and independent review pass after narrowing the test changes; existing child cleanup is unchanged.

Only these two handwritten examples and one existing handwritten test file change. The example fixes total ten added lines; generated files, SDK source, dependencies, and exports are untouched.

[chatgpt-codex-connector]

Codex Review Summary

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

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-09-05T20:29:57.744295Z	e819b59	PR opened
🔒 Security Review	✅ Completed 2026-09-05T20:30:47.220341Z	e819b59	PR opened

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.

[openai-sdks]

OkTest Summary

✅ 237/237 SDK tests passed in 10.462s for Node SDK PR #2645.

Test results — 42 files

Test	Result	Time
tests/chat-completions-complex-body.test.ts	✅ Passed	146ms
tests/chat-completions-create.test.ts	✅ Passed	245ms
tests/chat-completions-stream.test.ts	✅ Passed	210ms
tests/files-content-binary.test.ts	✅ Passed	171ms
tests/files-create-multipart.test.ts	✅ Passed	159ms
tests/files-list-pagination.test.ts	✅ Passed	163ms
tests/initialize-config.test.ts	✅ Passed	153ms
tests/instance-isolation.test.ts	✅ Passed	95ms
tests/models-list.test.ts	✅ Passed	130ms
tests/responses-background-lifecycle.test.ts	✅ Passed	191ms
tests/responses-body-method-errors.test.ts	✅ Passed	302ms
tests/responses-cancel-timeout.test.ts	✅ Passed	189ms
tests/responses-cancel.test.ts	✅ Passed	196ms
tests/responses-compact-retries.test.ts	✅ Passed	259ms
tests/responses-compact.test.ts	✅ Passed	201ms
tests/responses-create-advanced-stream.test.ts	✅ Passed	122ms
tests/responses-create-advanced.test.ts	✅ Passed	191ms
tests/responses-create-disconnect.test.ts	✅ Passed	1.095s
tests/responses-create-errors.test.ts	✅ Passed	204ms
tests/responses-create-malformed-api-responses.test.ts	✅ Passed	154ms
tests/responses-create-retries.test.ts	✅ Passed	256ms
tests/responses-create-stream-failures.test.ts	✅ Passed	217ms
tests/responses-create-stream-timeout.test.ts	✅ Passed	2.16s
tests/responses-create-stream-wire.test.ts	✅ Passed	2.624s
tests/responses-create-stream.test.ts	✅ Passed	124ms
tests/responses-create-terminal-states.test.ts	✅ Passed	313ms
tests/responses-create-timeout.test.ts	✅ Passed	188ms
tests/responses-create.test.ts	✅ Passed	209ms
tests/responses-delete.test.ts	✅ Passed	239ms
tests/responses-input-items-errors.test.ts	✅ Passed	243ms
tests/responses-input-items-list.test.ts	✅ Passed	193ms
tests/responses-input-items-options.test.ts	✅ Passed	107ms
tests/responses-input-tokens-count-timeout.test.ts	✅ Passed	211ms
tests/responses-input-tokens-count.test.ts	✅ Passed	336ms
tests/responses-malformed-inputs.test.ts	✅ Passed	2.28s
tests/responses-not-found-errors.test.ts	✅ Passed	241ms
tests/responses-parse.test.ts	✅ Passed	174ms
tests/responses-retrieve-retries.test.ts	✅ Passed	228ms
tests/responses-retrieve.test.ts	✅ Passed	298ms
tests/responses-stored-method-errors.test.ts	✅ Passed	591ms
tests/retry-behavior.test.ts	✅ Passed	2.957s
tests/sdk-error-shape.test.ts	✅ Passed	279ms

View OkTest run #33990207026

_{SDK merge (0e3cb8ad4e35) · head (e819b5908859) · base (7db48ecca499) · OkTest (2b1bdfd25e98)}

[github-actions]

Castiron custom code

✅ No new custom-code files detected.

32 mixed files remain; 0 existing customizations changed.

Compared 7db48ecca499 → e819b5908859. Generated baselines verified.

32 existing customizations unchanged

• api.md
• scripts/castiron/README.md
• scripts/castiron/custom_code_report.py
• scripts/castiron/test_custom_code_report.py
• src/client.ts
• src/resources/audio/transcriptions.ts
• src/resources/audio/translations.ts
• src/resources/beta/assistants.ts
• src/resources/beta/beta.ts
• src/resources/beta/index.ts
• src/resources/beta/responses/internal-base.ts
• src/resources/beta/responses/responses.ts
• src/resources/beta/threads/index.ts
• src/resources/beta/threads/runs/index.ts
• src/resources/beta/threads/runs/runs.ts
• src/resources/beta/threads/threads.ts
• src/resources/chat/completions/completions.ts
• src/resources/chat/completions/index.ts
• src/resources/conversations/index.ts
• src/resources/embeddings.ts
• src/resources/files.ts
• src/resources/fine-tuning/checkpoints/permissions.ts
• src/resources/images.ts
• src/resources/responses/internal-base.ts
• src/resources/responses/responses.ts
• src/resources/skills/skills.ts
• src/resources/skills/versions/versions.ts
• src/resources/vector-stores/file-batches.ts
• src/resources/vector-stores/files.ts
• src/resources/webhooks/index.ts
• src/resources/webhooks/webhooks.ts
• tests/lib/data-residency.test.ts

A changed generated baseline means this report cannot reliably identify which handwritten lines changed.

Inspect the custom-code diff

Download the exact patch produced by this run (requires repository access):

gh run download 33990226284 --repo openai/openai-node \
  --name castiron-custom-code-33990226284-1 --dir /tmp/castiron-custom-code-33990226284-1
git apply --stat /tmp/castiron-custom-code-33990226284-1/custom-code.patch
cat /tmp/castiron-custom-code-33990226284-1/custom-code.patch

Or reproduce it from an SDK checkout containing the vendored reporter:

git fetch --no-tags origin 7db48ecca4999a561faa60bb5040bf145d6e5684 e819b5908859e46cab83b226a0301a6e3a6e3ad5
python3 scripts/castiron/custom_code_report.py report \
  --base 7db48ecca4999a561faa60bb5040bf145d6e5684 \
  --head e819b5908859e46cab83b226a0301a6e3a6e3ad5 --fetch --require-head-hash --public \
  --out /tmp/castiron-custom-code-e819b5908859
cat /tmp/castiron-custom-code-e819b5908859/custom-code.patch

This is the current full custom patch for mixed files, not an attribution of only the handwritten lines changed by this PR.

Full report and patch

[jbeckwith-oai]

Reviewed the exact three-file head and traced finalResponse semantics for both real example entrypoints. Draining a stream is not treated as success: completed stays zero, while failed/incomplete and clean EOF with the latest in-progress snapshot emit a status-only diagnostic and set a nonzero exit. Existing event/final-response stdout, request parameters, strict query tool schema, normal request rejection, and SDK finalResponse behavior remain unchanged. The loopback integration adds both examples across completed/failed/incomplete/in-progress cases and verifies request count, auth, model/tool payloads, and private-detail non-leakage. Review threads are clear and CI is green. No blocking findings.

[markstuart-oai]

Reviewed all three changed files against ResponseStream, the accumulator/parser and iterator error contracts, and the streaming guide. The final status checks correctly distinguish completed responses from failed, incomplete and premature clean EOF snapshots, while preserving request/transport rejection. Keeping this small policy at each standalone example is clearer than introducing another helper; the existing loopback harness is reused without changing the SDK contract. No substantive findings.

Local validation: 55 ResponseStream/ResponsesParser tests passed, 32 synthetic base/head entrypoint cases passed, and canonical lint/format checks passed. The 34 loopback integration cases could not receive requests in this environment (a separate child-process probe reproduced request cancellation before server receipt). Full typechecking was limited by missing optional Azure/Express/dotenv example dependencies. I did not independently rerun the Node-version matrix.