Skip to content

Fix managed restart custody and readiness adoption - #292

Merged
adrianwebb merged 2 commits into
stagingfrom
codex/sdk-restart-readiness274
Oct 9, 2026
Merged

adrianwebb merged 2 commits into
stagingfrom
codex/sdk-restart-readiness274

Conversation

@adrianwebb

Copy link
Copy Markdown
Contributor

Refs #274; SDK #376.

Refresh the current manifest runtime before a non-plan restart, and use the existing manager-custody endpoint adoption path without redundant host-port probes. Preserve read-only plan, exact selection mode, failed refresh/build history and caller-owned port readiness.

Adds two unit regressions and one real native Git/caller-build integration regression with original golden scene/verifier bindings. Focused originals: 25/25 pass following 3 reproduced failures. Initial complete run failed on dirty index custody and missing native metadata allowlist; retained and corrected, not waived. Fresh complete suites and coded components must pass here before merge.

No new schema, runner, dependency, privilege, deadline, tag, UI or deployment. Temporary remote branch will be deleted after checked staging→main merges.

@adrianwebb
adrianwebb merged commit 63c7928 into staging Oct 9, 2026
1 of 2 checks passed
@adrianwebb
adrianwebb deleted the codex/sdk-restart-readiness274 branch October 9, 2026 15:09
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