Skip to content

fix: unblock structured and multi-group Codex reviews - #50

Merged
AppSolves merged 4 commits into
mainfrom
fix/structured-output-schema-contract
Nov 16, 2026
Merged

AppSolves merged 4 commits into
mainfrom
fix/structured-output-schema-contract

Conversation

@AppSolves

@AppSolves AppSolves commented Nov 16, 2026 •

Copy link
Copy Markdown
Owner

CodexBot could reject the first Responses request, demand redundant shell inspection in later review turns, or lose required image inputs after native Codex compaction. This fixes those three execution failures without changing inference budgets, credentials, authorization, quotas or approval/merge policy.

  • Project group/continuation/synthesis output schemas onto the documented Structured Outputs subset. Keep uniqueness and string-length constraints in strict host Ajv validation. Retain only exact allowlisted invalid_json_schema / text.format.schema diagnostics from bounded error envelopes.
  • Track successful tool inspection across the isolated execution. Every group's supplied evidence remains independently immutable/hash-verified; no-inspection, missing context, invalid output and unfinished synthesis still fail closed.
  • Preserve the first verified native image parts and generated labels in the per-execution proxy. Restore them only when compaction removes all images, then enforce the original fingerprints and byte/request/model limits. Partial loss, substituted images, duplicate-as-replacement and cross-execution reuse remain rejected.

Real ChatGPT acceptance on the exact final source 2a5f0e8:

Fixture Result Actual evidence
Small: 2 files / 2 commits Complete clean review, 1/1 group, zero findings Successful workflow, native review
Medium: 33 files Complete 3/3 groups and synthesis; expected cross-file authorization defect and correct inline finding Successful workflow, native review
Large: 93 files / 2 commits / 89 text patches / 4 distinct PNGs / 8,633 added lines Complete 6/6 groups and synthesis; zero findings Successful workflow, native review

All native events were COMMENT because the unchanged TopkAPI policy disables automated approval and requested changes. All results bind to the exact fixture heads. The staging webhook deliveries were operator-signed fixtures derived from fenced real comments, followed by actual App dispatch, Actions OIDC, protected Environment credentials, isolated Codex, official Responses inference and native publication. This is not evidence of direct GitHub webhook delivery to the separate test backend. No API-key fallback or KMU review was used.

Deterministic/native regressions reproduce the schema rejection, per-turn inspection failure and actual pinned-Codex image-dropping compaction before their corrections. The compaction case retains all four distinct images on every upstream request through six groups, continuation and synthesis. Validation passed: npm run check (246 tests, typecheck, lint, build); clean checkout with npm ci --ignore-scripts; workerd/D1; browser tests with synthetic providers; catalog; npm audit (zero vulnerabilities); Actionlint; Wrangler dry-run; Docker and all 19 pinned-Codex synthetic-provider inspection scenarios. Both full GitHub CI runs on this head passed, including image scanning and sandbox checks.

Cleanup completed: the official CLI confirmed revocation of the temporary renewable ChatGPT session; TopkAPI's original vault secret was restored securely through stdin. Cleanup PR29 restored the exact original workflow bytes and entire original main Git tree. Fixture PRs23–25 were closed without merging. The temporary Worker, D1 and local secret copy were deleted; resource absence and unchanged production deployment/health were independently checked. Production/KMU configuration and the production trusted workflow pin were not changed. This source PR remains unmerged.

Limitations: the historical production HTTP400 did not retain a usable provider rejection code, so its sole rejected keyword is not retrospectively proven. Successful live requests now establish the corrected path works for these fixtures, not arbitrary PRs or guaranteed model accuracy. An earlier small test's Unicode-escape false positive was disproved with the original bytes and isolated assertions; that review remains visible. The equivalent literal-input fixture passed, and the false positive is not concealed or counted as a clean success. Four small distinct PNGs establish image transport/provenance, not semantic accuracy on complex screenshots. All existing hard resource and completeness limits remain enforced.

@AppSolves AppSolves changed the title fix: keep host-only constraints out of Responses schemas fix: unblock Responses schemas and final review synthesis Nov 16, 2026
@AppSolves AppSolves changed the title fix: unblock Responses schemas and final review synthesis fix: unblock structured and multi-group Codex reviews Nov 16, 2026
@AppSolves
AppSolves merged commit a2b4073 into main Nov 16, 2026
4 checks passed
@AppSolves
AppSolves deleted the fix/structured-output-schema-contract branch November 16, 2026 02:27
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant