chore: willboosterify this repo#18
Conversation
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request updates testing protocols and coding style guidelines across several documentation files, replaces AI-specific check scripts with new verification commands, and adds the @willbooster/wb dependency. The feedback identifies opportunities to improve maintainability by consolidating duplicated rule files and reducing redundancy by removing obsolete scripts from the package configuration.
| - Continue modifying tests and/or code until all tests pass. | ||
| - Ensure tests are idempotent and independent (e.g., reset persistent data) so they can run repeatedly or in parallel. | ||
| - Prefer actual API calls over mocks, unless actual calls are impractical, have unintended side effects, or mocks are explicitly requested. | ||
| - Always investigate the root cause of a test failure before fixing it. | ||
| - Avoid adding wait functions in E2E tests unless you are confident they are necessary. | ||
| - Avoid adding wait functions in E2E tests unless strictly necessary. | ||
| - When fixing issues, follow these rules: | ||
| - Investigate the root cause first (e.g., by gathering debug logs, taking screenshots, etc.). | ||
| - Fix the actual root cause instead of applying workarounds. | ||
| - After making code changes, run `yarn check-all-for-ai` to execute all tests (takes up to 1 hour), or `yarn check-for-ai` for only type checking and linting (takes up to 10 minutes). | ||
| - If you are confident that your changes will not break any tests, you may use `check-for-ai`. | ||
| - After making code changes, run `yarn verify-full` to execute all tests (takes up to 1 hour), or `yarn verify` for only type checking and linting (takes up to 10 minutes). | ||
| - If you are confident that your changes will not break any tests, you may use `verify`. | ||
| - Use `oxlint` ignore comments with reasons (e.g., `// oxlint-disable-next-line <rule> -- <reason>`) if lint errors or warnings cannot be fixed. | ||
| - Once you have verified your changes, commit and push them to the current (non-main) branch, then create a PR via `gh`. | ||
| - Follow the conventional commits; your commit message should start with `feat:`, `fix:`, etc. | ||
| - If not specified, make sure to add a new line at the end of your commit message with: `Co-authored-by: WillBooster (Codex CLI) <agent@willbooster.com>`. | ||
| - Always create new commits. Avoid using `--amend`. | ||
| - Always use heredoc syntax when passing multi-line content to any command. | ||
| - Use `wb start --mode test` to launch a web server for debugging or testing. |
There was a problem hiding this comment.
There is significant content duplication across .cursor/rules/general.mdc, AGENTS.md, CLAUDE.md, and GEMINI.md. This pull request modifies all four files with nearly identical changes, which highlights a maintainability issue.
To improve maintainability and reduce redundancy, consider consolidating these files. You could have a single source of truth for the common rules and only specify the differences (like the Co-authored-by line) in separate, smaller files if needed. This would make future updates much easier.
References
- Simplify code as much as possible to eliminate redundancy. The presence of four nearly identical files for agent rules violates this principle. (link)
| "verify": "wb verify", | ||
| "verify-full": "wb verify --full" |
There was a problem hiding this comment.
The new verify and verify-full scripts are added, and the documentation files are updated to refer to them. However, the old scripts check-all-for-ai and check-for-ai are still present in package.json but are no longer mentioned in the updated documentation.
To avoid confusion and to align with the style guide's principle of eliminating redundancy, consider removing the old check-all-for-ai and check-for-ai scripts.
References
- Simplify code as much as possible to eliminate redundancy. Keeping old, unused scripts in
package.jsonis a form of redundancy. (link)
1 participant
No description provided.