Add Shipyard CI workflow and PR template - #40
Conversation
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
Codex Review: Here are some suggestions.
Reply with @codex fix comments to fix any unresolved comments.
About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you open a pull request for review, mark a draft as ready, or comment "@codex review". If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex fix this CI failure" or "@codex address that feedback".
| - name: Install dependencies | ||
| run: pnpm i --frozen-lockfile |
There was a problem hiding this comment.
[P0] Use npm install to match lockfile
The workflow installs dependencies via pnpm i --frozen-lockfile, but the repo only ships package-lock.json and no pnpm-lock.yaml. pnpm aborts with “No lockfile was found” when --frozen-lockfile is set, so the job will fail before typecheck, lint, or tests can run. Switch the install step to npm ci (or commit a pnpm lockfile) so CI can complete.
Useful? React with 👍 / 👎.
| - name: Typecheck | ||
| run: pnpm typecheck |
There was a problem hiding this comment.
[P0] Define typecheck script before calling it
The Typecheck step invokes pnpm typecheck, but package.json declares only dev, build, start, and lint scripts. GitHub Actions will exit with “Missing script: "typecheck"” and the workflow stops at this step. Add a typecheck script (e.g. tsc --noEmit) or call an existing command so the step can succeed.
Useful? React with 👍 / 👎.
| - name: Test | ||
| run: pnpm test -- --reporter=dot || true |
There was a problem hiding this comment.
[P1] Ensure test failures block CI
The Test step runs pnpm test -- --reporter=dot || true, but no test script exists in package.json and the trailing || true forces the step to succeed even if the command errors out. As a result, tests never execute and their failures cannot fail the workflow, so regressions will be missed. Define a real test script and let the step fail when tests fail.
Useful? React with 👍 / 👎.
1 participant
Summary
Testing
https://chatgpt.com/codex/tasks/task_e_68c852169e348333b8f933efcf923ab0