mirror of
https://github.com/max-sixty/worktrunk.git
synced 2026-09-14 20:00:38 +08:00
Worktrunk Bot
5f06e55332
## Summary Weekly CI renovation. One cargo-install pin had drifted; everything else in the weekly checklist is already current. Rust stable is 1.98.0, so MSRV and the development toolchain stay at 1.97 (latest stable − 1) — last week's bump already landed them there. - **`worktrunk`: 0.74.0 → 0.75.0** — `ci.yaml` ×2, `nightly.yaml` ×1. `rust-version` 1.97, exactly the pinned toolchain. Toolchain compatibility for every pinned crate against 1.97.0: `cargo-affected` 0.4.0 (1.94), `cargo-insta` 1.48.0 (1.66.0), `cargo-nextest` 0.9.143 (1.91), `cargo-llvm-cov` 0.9.0 (1.87), `cargo-msrv` 0.19.3 (1.91.1), `lychee` 0.24.2 (1.88.0), `worktrunk` 0.75.0 (1.97), `cargo-udeps` 0.1.61 (unspecified). All build under 1.97.0. <details><summary>Everything else the weekly pass checked, and why it isn't here</summary> - **MSRV + toolchain** — stable is 1.98.0 (released 2026-08-18), so the target is 1.97. `Cargo.toml`, `tests/helpers/wt-perf/Cargo.toml`, and `rust-toolchain.toml` are all already at 1.97. No change, so no `flake.lock` refresh either. - **Pinned nightlies** — `nightly-2026-08-01` ×2 in `nightly.yaml`. One month old; the rule bumps only past three months. - **Other cargo-install pins** — `cargo-affected`, `cargo-insta`, `cargo-nextest`, `cargo-llvm-cov`, `cargo-msrv`, `cargo-udeps`, `lychee` all match their current crates.io release. - **Codex Cloud / `setup-web`** — `pre-commit` 4.6.2, `cargo-insta` 1.48.0, `cargo-nextest` 0.9.143, PowerShell 7.6.5 all current. Nushell has drifted and is handled separately, below. - **`ubuntu-24.04`** — still what `ubuntu-latest` resolves to, so it stays. Ubuntu 26.04 is badged `preview`, which is not a bump target. - **Dependabot's `typescript` major ignore** — stays. `@astrojs/check` still declares `peer typescript@"^5.0.0 || ^6.0.0"` and the current `typescript` major is 7, so the entry is still the only thing keeping a TypeScript 7 lockfile from breaking `npm ci`. - **LLM model pins in docs** — `claude-haiku-4.5` and `gpt-5.6-luna` are both still the fastest/smallest in their vendor's current family. OpenAI's models page describes Luna as "fast and affordable … at the lowest cost in the family"; the only smaller Codex model, `gpt-5.3-codex-spark`, is a Pro-only research preview and not a good default recommendation. - **Statusline cache-check** — clean. `wt config state logs profile` reports `same_context_duplicates: []` and 0 extra calls on a `wt list statusline --format=claude-code` render. </details> Three findings from this pass are **not** in this PR, because each can turn a whole CI leg red and belongs on its own branch: the Nushell bump (moves PTY snapshots) and the two runner-image bumps (`macos-15` and `windows-2022` have both fallen off their `-latest` label). Co-authored-by: worktrunk-bot <254187624+worktrunk-bot@users.noreply.github.com>