chore(release): bump dev version to 0.1.57 after 0.1.56 release

[andrei-hasna]

Summary

Reconciles the release-line drift that blocked release automation for hasna/codewith. The 0.1.56 release is already fully materialized at HEAD 3479752353fa:

• tag rust-v0.1.56 dereferences exactly to HEAD
• npm @hasna/codewith@0.1.56 already published
• GitHub release rust-v0.1.56 already exists

main still declared the already-published 0.1.56, which is the root cause of the tag/npm/release "already exists" blockers on the next release line. This PR advances the in-repo dev version so the next release no longer collides. There are no unreleased functional commits beyond the 0.1.56 tag, so a re-release would be wrong — advancing the declared version is the correct reconciliation.

Changes

• codex-rs/Cargo.toml [workspace.package].version: 0.1.56 → 0.1.57
• codex-cli/package.json version: 0.1.56 → 0.1.57 (keeps npm package in lockstep with the rust workspace)
• codex-rs/Cargo.lock: 127 workspace-crate entries regenerated to 0.1.57 (via cargo update --workspace --offline; only workspace crates changed — the unrelated external cmake@0.1.57 entry is untouched)
• MODULE.bazel.lock: regenerated Bazel module facts for the updated Cargo lock so Bazel CI lockfile checks match the Cargo version bump

Not touched: sdk/typescript/package.json (independent 0.1.0), root package.json (no version field), CHANGELOG.md ([Unreleased] remains the accumulation point; no dated 0.1.57 section until an actual release).

Guardrails honored

No tag create/push, no npm/bun publish, no release-workflow dispatch. This is a version-declaration bump only.

Validation

• cargo metadata --no-deps --format-version=1 --offline → OK (manifests parse, lock consistent)
• python3 .github/scripts/verify_cargo_workspace_manifests.py → OK
• grep -rn '0\.1\.56' codex-rs/Cargo.toml codex-cli/package.json codex-rs/Cargo.lock → 0 matches
• Cargo.lock diff: exactly 127 insertions / 127 deletions, all 0.1.56 → 0.1.57 version bumps
• bazel --batch ... mod deps --lockfile_mode=error with repo/disk caches redirected under /tmp → OK (just bazel-lock-check could not run unmodified in this sandbox because .bazelrc points Bazel caches at read-only ~/.cache)
• CARGO_TARGET_DIR=/tmp/pr117-cargo-target just check-fast -p codex-cli → OK

Note: 5 open PRs (#116, #115, #114, #113, #93) are future release content and are intentionally not touched by this task.

🤖 Generated with Claude Code

---

Note

Low Risk
Version-only bump with lockfile regeneration; no application logic or release artifacts are modified.

Overview
Advances the in-repo development version from 0.1.56 to 0.1.57 so release automation no longer collides with the already-shipped 0.1.56 tag and npm publish.

codex-rs/Cargo.toml [workspace.package].version and codex-cli/package.json are bumped together; codex-rs/Cargo.lock updates all workspace crate versions to 0.1.57. MODULE.bazel.lock picks up minor registry metadata (e.g. anyhow_1.0.103, tar_0.4.45). No functional or changelog changes—version declaration only.

^{Reviewed by Cursor Bugbot for commit 28d5fad. Bugbot is set up for automated code reviews on this repo. Configure here.}

[andrei-hasna]

[codewith] OpenLoops worker update for task 6887edb4-dae4-42ed-9630-22d5003e5f4e.

Pushed amended head 28d5fad58d1a01834640ef918c4b0ee36e589cc1 with:

• MODULE.bazel.lock regenerated for the Cargo version bump, fixing the branch-related Bazel lockfile failure seen on 98abc5ed799aefb705af69d5cba987400cea15d4
• commit body amended to remove the banned Co-Authored-By trailer
• PR body updated to include the Bazel lockfile change and current validation

Validation passed locally on the new head:

• cargo metadata --no-deps --format-version=1 --offline
• python3 .github/scripts/verify_cargo_workspace_manifests.py
• no 0.1.56 version declarations remain in the changed version files
• Bazel 9 mod deps --lockfile_mode=error passed with repo/disk caches redirected under /tmp for this sandbox
• CARGO_TARGET_DIR=/tmp/pr117-cargo-target just check-fast -p codex-cli
• staged secrets scan before commit amend

Not merged yet: GitHub currently reports mergeable=MERGEABLE but mergeStateStatus=UNSTABLE because the new check suite is still pending. Review request to selected non-author reviewer kriptoburak failed with HTTP 422 because GitHub says that user is not a collaborator on hasna/codewith.

[andrei-hasna]

[codewith] OpenLoops post-push gate update for task bbc4b907-df34-497b-b5df-c9fad1968b61.

Not merging PR #117 at head 28d5fad58d1a01834640ef918c4b0ee36e589cc1.

Current gate state:

• PR is OPEN and non-draft; base main is 3479752353fabb9cd933992a23c936a8aa3e8319.
• GitHub reports mergeable=MERGEABLE but mergeStateStatus=UNSTABLE.
• Check rollup is still blocked: 5 failed, 12 pending, 23 passed.
• Review state is not satisfied: reviewDecision is empty, no latest reviews, and no review requests. The earlier request to selected non-author reviewer kriptoburak failed with GitHub HTTP 422 because review requests require collaborators.
• Commit-message policy is clean for the current head: no Co-Authored-By trailer found.
• Branch freshness/conflict check is clean: origin/main...28d5fad is 0 behind / 1 ahead, and git merge-tree origin/main 28d5fad... produced no conflict markers.

Failed CI blockers diagnosed from run 28728169869:

• Linux Bazel job 85189185251: //codex-rs/state:state-unit-tests-bin fails to compile because Bazel/rustc cannot read codex-rs/state/src/runtime/fixtures/goals_0148_0005_thread_goal_deferred.sql.
• Windows Bazel shard 1 job 85189185249: //codex-rs/codex-api:codex-api-unit-tests fails endpoint::models::tests::parses_nvidia_models_response; //codex-rs/models-manager:models-manager-unit-tests fails model_info::tests::known_nvidia_deepseek_model_uses_local_metadata.
• Windows Bazel shard 2 job 85189185254: app-server unit/all tests fail provider fallback, external-agent env, schedule/thread, and v2 model-list cases.
• Windows Bazel shard 3 job 85189185276: workflows OpenRouter tests, state 0148 tests, app-server background-agent tests, and TUI snapshot tests fail.
• Aggregate Windows Bazel job 85192198635 fails because the shard set failed.

Follow-up todo routing created/updated for the blockers:

• b9ad1d75: existing workflows OpenRouter Windows routing blocker / PR fix(workflows): make routing fixture injection CRLF-safe #119.
• 6993c8c3: updated existing TUI snapshot blocker with current PR chore(release): bump dev version to 0.1.57 after 0.1.56 release #117/main evidence and auto-route metadata.
• 0df33de8: new state/Bazel 0148 runtime fixture blocker.
• b26896cd: new provider model metadata/fallback blocker.
• 0814a6ed: new app-server background-agent/thread Windows Bazel blocker.

Main-branch comparison: latest main Bazel run 28708512474 at 3479752353fabb9cd933992a23c936a8aa3e8319 already fails matching Windows provider/app-server/workflows/state/TUI tests. The Linux state fixture compile failure is newly exposed after this PR's lockfile update gets Linux Bazel past the prior lock check, but the missing Bazel compile data is outside PR #117's changed files.

[andrei-hasna]

Closing as obsolete after live PR0 wave2 recheck. This was a 0.1.57 dev-version bump; main is now 0.1.60 (codex-cli package plus Cargo.toml), and the PR is conflicted with failing required checks. No merge path remains.