[AdaWorldAPI]

Both jobs added by the previous PR fail before compiling anything:

error: failed to load source for dependency `ndarray`
Caused by: unable to update /home/runner/work/a2ui-rs/ndarray
Caused by: No such file or directory (os error 2)

.cargo/config.toml carries

[patch."https://github.com/AdaWorldAPI/ndarray"]
ndarray = { path = "../ndarray" }

so every cargo invocation in this repository redirects the ndarray git coordinate to a sibling checkout — cargo fmt included, because it resolves the workspace through cargo metadata.

That file describes itself as a local development override the shipped build never sees, and it is right about the Railway image, whose Dockerfile copies only the manifests and crates/. It is wrong about CI: the file is committed, so a runner checks it out and cargo honours it. Nothing had ever run cargo in this repository before this week, so the assumption went untested until a gate arrived to test it.

The fix

Deleting the override to suit CI would discard a deliberate operator ruling — prefer the local fork source over a fetch — so the jobs give CI the sibling instead. actions/checkout cannot write outside $GITHUB_WORKSPACE, so this repository is checked out into a named subdirectory with ndarray beside it. That is the same shape AdaWorldAPI/lance-graph's own style.yml already uses for its siblings.

Each setup-rust-toolchain also gains rust-src-dir, which is not incidental. defaults.run.working-directory governs run: steps only, so without it the action reads no rust-toolchain.toml and installs stable instead of the pin — the identical trap this sweep hit in lance-graph-java, where a test caught it.

Verification

Run in the layout CI now reproduces: cargo metadata resolves, which is precisely the step that was failing, and cargo fmt --all --check is clean. Clippy and the test suite are CI's to measure — the workspace pulls a wgpu fork and is larger than this machine can build, so those two jobs get their first real measurement here.

---

Generated by Claude Code

https://claude.ai/code/session_01AGVLyRZNEKKBSfBDJfbY3V

Summary by CodeRabbit

• Chores
  • Updated the lint workflow to run formatting, Clippy checks, and tests from the designated project checkout.
  • Improved Rust toolchain setup and cache configuration for more reliable continuous integration checks.

[coderabbitai]

Caution

Review failed

The pull request is closed.

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Essentials

Run ID: b90156f0-9443-4621-8901-9234da5aa9c5

📥 Commits

Reviewing files that changed from the base of the PR and between edf6174 and 6121450.

📒 Files selected for processing (1)

• .github/workflows/lint.yml

---

📝 Walkthrough

Walkthrough

The lint workflow now runs format, Clippy, and test jobs from a nested a2ui-rs checkout. Each job also checks out the sibling ndarray repository and uses the nested workspace and Rust toolchain.

Changes

Lint workflow setup

Layer / File(s)	Summary
Nested checkout validation jobs .github/workflows/lint.yml	Format, Clippy, and test jobs use the nested a2ui-rs checkout, sibling ndarray checkout, nested Rust toolchain, and scoped workspace cache.

Estimated code review effort: 2 (Simple) | ~10 minutes

Suggested reviewers: claude

Poem

I hop through workflows, tidy and bright
Nested paths guide each check just right
Clippy and tests march in a row
With ndarray beside them, ready to go
Format leaves every line aglow

---

_{Comment @coderabbitai help to get the list of available commands.}

[cursor]

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: serverGenReqId_66931193-08f9-4efa-b59d-eedcff05957d)

[chatgpt-codex-connector]

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.
To continue using code reviews, you can upgrade your account or add credits to your account and enable them for code reviews in your settings.