[dpiet-oai]

Summary

Use the active authenticated main policy when evaluating the initial next policy pin.

Changes

• Keep the policy selector and review-status invalidation workflow aligned during initial pin adoption.
• Retain exact-head human attestation for effective policy relaxation, checker rollback, local budget increases, and ratchet activation.
• Cover bootstrap policy decisions and review-status invalidation with offline regression checks.

[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 2026-10-23T03:15:12.083410Z	51ba8fa	PR opened
🔒 Security Review	✅ Completed 2026-10-23T03:16:26.342752Z	51ba8fa	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.

[github-actions]

Castiron custom code

Evaluated main: d8653786396518d29eaeb9ec112d5209ed148489.

Trusted main policy: d8653786396518d29eaeb9ec112d5209ed148489.

✅ No new custom-code files detected.

0 mixed files remain; 0 existing customizations changed.

Compared d86537863965 → 51ba8fae0277. Generated baselines verified.

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 39893945856 --repo openai/openai-java \
  --name castiron-custom-code-39893945856-1 --dir /tmp/castiron-custom-code-39893945856-1
git apply --stat /tmp/castiron-custom-code-39893945856-1/custom-code.patch
cat /tmp/castiron-custom-code-39893945856-1/custom-code.patch

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

git fetch --no-tags origin d8653786396518d29eaeb9ec112d5209ed148489 51ba8fae02773e2bd11b899dfccb8b0861157d54
python3 scripts/castiron/custom_code_report.py report \
  --base d8653786396518d29eaeb9ec112d5209ed148489 \
  --head 51ba8fae02773e2bd11b899dfccb8b0861157d54 --target-branch main --fetch --require-head-hash --public \
  --out /tmp/castiron-custom-code-51ba8fae0277
cat /tmp/castiron-custom-code-51ba8fae0277/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

[dpiet-oai]

Prerequisite for #1268: keeps the generated policy/workflow updates separate from the Decisions SDK changes. During initial next pin adoption, both the selector and review-status invalidation compare against the active authenticated main policy. Effective policy relaxations, checker rollbacks, local budget increases, and ratchet activation still require exact-head human approval.

Validation: strict generated-ownership verification passed; 231 policy tests ran (1 skipped). No SDK models or API schema/config changed.

[markstuart-oai]

Reviewed the Java policy source at 51ba8fae02773e2bd11b899dfccb8b0861157d54. Reviewed initial next-policy adoption and review-change invalidation. Both now compare the proposed effective policy with authenticated main during bootstrap; existing pins still compare with the protected target. Rollbacks, budget increases and mode changes retain exact-head human approval checks. The regression cases cover PR and queue selection plus invalidation after main advances. No actionable source finding.

Static source review only; I did not run candidate code, tests, or builds.

This automated source review is not human policy approval. Existing Java #1268 feedback remains unchanged.

Current hosted checks report success or skipped.