Skip to content

CONTRIBUTING.md

Latest commit

 

History

History
64 lines (43 loc) · 2.4 KB

CONTRIBUTING.md

File metadata and controls

64 lines (43 loc) · 2.4 KB

Contributing

Thanks for wanting to contribute! Bug reports, feature requests, and PRs are all welcome.

If you're planning something non-trivial, open an issue first so we can all talk through it before you spend time on a PR.

Setup

git clone https://github.com/mcpcat/webmcp-react.git
cd webmcp-react
pnpm install

You'll need pnpm since the repo uses pnpm workspaces.

Development

Command What it does
pnpm build Build CJS + ESM + types via tsup
pnpm test Run tests via Vitest
pnpm test:watch Tests in watch mode
pnpm typecheck TypeScript type check
pnpm lint Biome lint + format check
pnpm lint:fix Auto-fix lint + format issues
pnpm dev:playground Vite playground app
pnpm dev:nextjs Next.js example app

There are precommit hooks (Husky + lint-staged) that auto-format staged .ts, .tsx, and .json files with Biome when you commit. This gets set up automatically when you run pnpm install.

The playground and Next.js example are handy for testing changes manually since not everthing is easy to cover with unit tests.

Pull Requests

  • One concern per PR
  • Add tests for new stuff. Tests live in src/**/__tests__/
  • Before pushing, make sure everything passes:
pnpm build && pnpm typecheck && pnpm lint && pnpm test

Code Style

Biome handles formatting and linting, and the pre-commit hooks take care of it automatically, so you mostly don't need to worry about this. Run pnpm lint:fix if you want to fix things manually.

Commit messages: simple is preferred, but they get squashed anyway PR: useful titles and description (TBH you should use AI to generate them anyway)

AI-Assisted Contributions

This library literally exists to connect React apps with AI agents, so it would be pretty hypocritical to not encourage AI-assisted contributions. :)

If you used AI tools, just mention it brefly in your PR (what tool, how you used it). "Scaffolded tests with Claude, refined manually" that kind of thing is fine.

Otherwise the same rules apply as any other contribution:

  • Understand your code. If a reviewer asks why something works, "the AI wrote it" isn't a great answer.
  • Test it. Run the suite, check edge cases.
  • Review every changed line before submitting.

License

Contributions are licensed under MIT, same as the project.