Thanks for your interest in contributing to wagmi! Please take a moment to review this document before submitting a pull request.
If you want to contribute, but aren't sure where to start, you can create a new discussion.
Note
Please ask first before starting work on any significant new features. This includes things like adding new connectors, hooks, chains, API providers, etc.
It's never a fun experience to have your pull request declined after investing time and effort into a new feature. To avoid this from happening, we request that contributors create a feature request to first discuss any API changes or significant new ideas.
This guide is intended to help you get started with contributing. By following these steps, you will understand the development process and workflow.
- Cloning the repository
- Installing Node.js and pnpm
- Installing dependencies
- Starting the development playground
- Running the test suite
- Writing documentation
- Submitting a pull request
This guide covers more advanced topics. Pick the topics based on your needs.
To start contributing to the project, clone it to your local machine using git:
git clone https://github.com/wagmi-dev/wagmi.git --recurse-submodulesOr the GitHub CLI:
gh repo clone wagmi-dev/wagmi -- --recurse-submoduleswagmi uses pnpm workspaces to manage multiple projects. You need to install Node.js v16 or higher and pnpm v7 or higher.
You can run the following commands in your terminal to check your local Node.js and npm versions:
node -v
pnpm -vIf the versions are not correct or you don't have Node.js or pnpm installed, download and follow their setup instructions:
- Install Node.js using fnm or from the official website
- Install pnpm
Once in the project's root directory, run the following command to install the project's dependencies:
pnpm installAfter the install completes, pnpm links packages across the project for development and git hooks are set up.
To start the local development playground, run the following. This will run a Next.js app (located at examples/_dev) that is set up for playing around with code while making changes.
pnpm playgroundOnce the Next.js dev server is running, you can make changes to any of the package source files (e.g. packages/react) and it will automatically update the playground. (If the playground isn't automatically updating, try running pnpm dev to relink packages in development mode.)
wagmi uses Anvil to execute tests against a local Ethereum node. First, install Anvil via Foundry. Next, add the following to your environment (recommended to use direnv):
ANVIL_FORK_URL=https://eth-mainnet.alchemyapi.io/v2/<apiKey>ANVIL_FORK_URL can be for any RPC service provider (e.g. Alchemy or Infura). Now you are ready to run the tests! In one terminal session, spin up Anvil using pnpm anvil. Next, in a different terminal session, you have the following options for running tests:
pnpm test— runs tests in watch modepnpm test:run— performs single run without watch mode
When adding new features or fixing bugs, it's important to add test cases to cover the new/updated behavior. If snapshot tests fail, you can run the test:update command to update the snapshots.
Documentation is crucial to helping developers of all experience levels use wagmi. wagmi uses Nextra and MDX for the documentation site (located at docs). To start the site in dev mode, run:
pnpm docs:devTry to keep documentation brief and use plain language so folks of all experience levels can understand. If you think something is unclear or could be explained better, you are welcome to open a pull request.
When you're ready to submit a pull request, you can follow these naming conventions:
- Pull request titles use the Imperative Mood (e.g.,
Add something,Fix something). - Changesets use past tense verbs (e.g.,
Added something,Fixed something).
When you submit a pull request, GitHub will automatically lint, build, and test your changes. If you see an ❌, it's most likely a bug in your code. Please, inspect the logs through the GitHub UI to find the cause.
When adding new features or fixing bugs, we'll need to bump the package versions. We use Changesets to do this.
Note
Only changes to the codebase that affect the public API or existing behavior (e.g. bugs) need changesets.
Each changeset defines which package(s) should be published and whether the change should be a major/minor/patch release, as well as providing release notes that will be added to the changelog upon release.
To create a new changeset, run pnpm changeset. This will run the Changesets CLI, prompting you for details about the change. You’ll be able to edit the file after it’s created — don’t worry about getting everything perfect up front.
Since we’re currently in beta, all changes should be marked as a minor/patch release to keep us within the v0.x range.
Even though you can technically use any markdown formatting you like, headings should be avoided since each changeset will ultimately be nested within a bullet list. Instead, bold text should be used as section headings.
If your PR is making changes to an area that already has a changeset (e.g. there’s an existing changeset covering theme API changes but you’re making further changes to the same API), you should update the existing changeset in your PR rather than creating a new one.
The first time a PR with a changeset is merged after a release, a new PR will automatically be created called chore: version packages. Any subsequent PRs with changesets will automatically update this existing version packages PR. Merging this PR triggers the release process by publishing to npm and cleaning up the changeset files.
If a PR has changesets, you can create a snapshot release by manually dispatching the Snapshot workflow. This publishes a tagged version to npm with the PR branch name and timestamp.