[COMPATIBILITY] Check Windows text output and pagination

[saioai]

The CLI already had tests for resource output, pagination, and streamed text, but Windows CI did not run them.
Windows x64 and ARM64 now run those tests before changes merge.
For example, CI checks that pagination preserves item order and that streaming failures retain text already printed.

What changes

• Check resource summaries, pagination limits, explicit formats, unknown fields, and partial-page failures.
• Check streamed text delivery, event identity, explicit formats, failure diagnostics, and large events.
• Require 13 resource/navigation roots with 100 children, plus 10 streaming roots with 41 children.
• Reject missing tests, unexpected passing names, skips, failures, and missing package passes.

Commands

No CLI commands or flags change. The two CI selections run:

# Resource output and pagination
go test -p 2 ./cmd/openai -count=1 -timeout=3m -v -json -run '^(TestMainResourceSummariesRetainActionFields|TestMainResourceSummaryListFilesInVectorBatch|TestMainResourceSummaryPaginationPreservesOrderDuplicatesAndLimits|TestMainResourceSummaryListNoticeForMixedRecords|TestMainResourceSummaryEmptyList|TestMainResourceSummaryUnknownFieldsAndShapesFallBack|TestMainResourceSummaryExplicitFormatsAndExtractionKeepFullData|TestMainResourceSummaryListFormatsAndExtractionKeepFullItems|TestMainResourceSummaryLeavesOtherOperationsUnchanged|TestMainResourceSummaryPreservesLargeActionArguments|TestMainResourceSummaryPreservesPartialPageOnFailure|TestMainResourceSummaryWritesBeforeNextPageArrives|TestMainListNavigationNoninteractive)$'

# Streaming text output
go test -p 2 ./cmd/openai -count=1 -timeout=3m -v -json -run '^(TestMainStreamingTextArrivesBeforeCompletion|TestMainStreamingTextKeepsPartIdentityAndSnapshots|TestMainStreamingTextRetainsNontextDetails|TestMainStreamingExplicitFormatsPreserveEveryEvent|TestMainStreamingTextInterleavedChoicesKeepDetails|TestMainStreamingFailuresPreservePartialOutput|TestMainStreamingInterruptedTransportPreservesPartialOutput|TestMainStreamingFailureDiagnostics|TestMainStreamingTextDistinguishesEarlyEOFAndEmptyCompletion|TestMainStreamingTextLargeEvent)$'

The selectors exclude the explicit-limit streaming test and the already-covered resumed-response dispatch test.

Code

Only .github/workflows/help-compatibility.yml changes. The steps reuse the existing PowerShell wrapper and result verifier.
The original 18 MiB resource case and large stream fixtures remain unchanged and sequential.

PR #452 merged into main as 28cd2647. This now targets that shipping base with only the resource and streaming additions.

Tested

• YAML parsing and all 205 verifier controls passed again after merging main.
• Native Windows x64 and Windows ARM64 passed at 767f946. Both used SDK 3.78.0, Go 1.26.9, and -p 2.
• Each architecture passed 13 resource/list roots with 100 children, plus 10 streaming roots with 41 children.
• Every selected test ran and passed exactly once. Each suite recorded one package pass, without selected skips or failures.
• Independent review verified the raw results and every prepared source file against the candidate.

All six Help compatibility jobs passed. CodeQL and Castiron checks passed.
Codex code and security reviews completed for 767f946 with no reported findings.
Main CI passed on its first attempt at 767f946.

Earlier b8407cd1 results used SDK 3.77 and remain historical.
Parent #452's earlier image-test timeout and unchanged passing retry remain documented. The timeout's underlying cause remains unresolved.

For #389 and #390, this covers noninteractive preservation only.
Files, Batches, and Projects do not activate compact tables or interactive navigation.
Shared Models behavior and helper coverage require separate evidence.
These checks do not establish terminal appearance, keyboard behavior, WSL, shell parsing, or user-signal cancellation.
The existing POSIX scripts/run test skips on Windows, outside these selections.

Demo

No visual demo is needed for this CI-only change. CLI behavior and appearance stay unchanged.

[saioai]

@codex review

[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 2026-10-13T23:09:37.796924Z	767f946	New commits
🔒 Security Review	✅ Completed 2026-10-13T23:09:11.135568Z	767f946	New commits

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.

[chatgpt-codex-connector]

Codex Review: Didn't find any major issues. Can't wait for the next one!

Reviewed commit: b8407cd17f

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".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

🛡️ Codex Security Review · Automatically triggered

Security review completed. No security issues were found in this pull request.

Reviewed commit: b8407cd17f

View security finding report

Only the user who started this review can view the report in Codex.

ℹ️ About Codex security reviews in GitHub

This is an experimental Codex feature. Security reviews are triggered when:

• You comment "@codex security review"
• A regular code review gets triggered (for example, "@codex review" or when a PR is opened), and you’re opted in so security review runs alongside code review

Once complete, Codex will leave suggestions, or a comment if no findings are found.

[markstuart-oai]

Reviewed all 1 changed files at b8407cd17f85cdac7cc5a9b93a385598ee6de3fe. I found no blocking issue.

The Windows jobs add the existing resource and streaming-text tests. The required names match the source, including names with spaces, slashes, and JSON text. The verifier requires all expected results and package success. Application behavior and production code are unchanged.

All non-skipped checks on this head passed. I reviewed source and hosted results only; no local tests, builds, or native-shell execution ran.

[github-actions]

Castiron custom code

Evaluated main: 28cd2647d05ed5f831c9f2e69b2cb1c27f4bf356.

✅ No new custom-code files detected.

5 mixed files remain; 0 existing customizations changed.

Compared 28cd2647d05e → 767f9465b393. Generated baselines verified.

5 existing customizations unchanged

• pkg/cmd/adminorganizationcertificate.go
• pkg/cmd/audiovoice.go
• scripts/castiron/custom_code_budget.py
• scripts/castiron/custom_code_report.py
• scripts/castiron/test_custom_code_budget.py

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 38483787319 --repo openai/openai-cli \
  --name castiron-custom-code-38483787319-1 --dir /tmp/castiron-custom-code-38483787319-1
git apply --stat /tmp/castiron-custom-code-38483787319-1/custom-code.patch
cat /tmp/castiron-custom-code-38483787319-1/custom-code.patch

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

git fetch --no-tags origin 28cd2647d05ed5f831c9f2e69b2cb1c27f4bf356 767f9465b3937f0b5106d5d957e5dd1b2570a127
python3 scripts/castiron/custom_code_report.py report \
  --base 28cd2647d05ed5f831c9f2e69b2cb1c27f4bf356 \
  --head 767f9465b3937f0b5106d5d957e5dd1b2570a127 --fetch --require-head-hash --public \
  --out /tmp/castiron-custom-code-767f9465b393
cat /tmp/castiron-custom-code-767f9465b393/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