A CLI for finding issues labeled with Good First Issue to help lower the barrier to contributing to open source projects.

• Good First Issue
  • Table of Contents
  • Prerequisites
  • Usage
    • Installation
    • Commands
    • CLI Options
  • TODOs: What's coming up next
  • Projects
  • Adding New Projects
    • Adding New Projects: More Information
  • Release Process
    • Versioning
  • Labels and Milestones
    • Local Testing
  • Contributing

To use Good First Issue, you'll need to have a few things installed:

• Node.js 8.0.0 or above:
  • If you need to install Node.js, you can download it from the official downloads page.
  • If you want to use a dynamic version manager, you can use something like nvm on macOS, Linux, and WSL.
• npm 5.0.0 or above:
  • If you already have Node.js 8.0.0 or above, you will have npm 5.0.0 or above.
  • If you need to update your npm CLI, run npm i -g npm.

This module is an interactive CLI. If you're looking for a module to use in an application, check out libgfi.

The suggested usage is via npx:

npx good-first-issue [project] # temporarily install and run the module, optionally passing `project`

Alternatively, you could absolutely install good-first-issue as a global module:

npm i -g good-first-issue # install globally
good-first-issue # call the CLI

• good-first-issue: open up the interactive project selection tool.
• good-first-issue [project]: you can pass in a name from the list of projects which is a curated list of projects that have been verified to have good-first-issues.
• good-first-issue [GitHub organization or user]: similar to [project] but will search any GitHub organization or user that exists for issues labeled with "Good First Issue".
• good-first-issue [GitHub organization or user]/[repo]: similar to [project], but will search a specific repository on GitHub within the organization for issues labeled with "Good First Issue".

• -o, --open - open in browser
• -f, --first - Return first/top issue
• -a, --auth <github personal access token> - Authenticate with the GitHub API (increased rate limits)

good-first-issue is still in an early state. I wanted to get good-first-issue node out the door, but have some other things I'm planning on implementing. Here's a list:

• good-first-issue node command
• Interactive selector when good-first-issue is run without a sub command
• Add additional useful commands
• Explore adding a secondary selector that shows paginated results from GitHub, allowing the user to select which Good First Issue to pick rather than returning a random one
• Improve Feeling Lucky to be better about picking a random issue
• Add more tests

If you'd like to help with any of these, feel free to submit a PR or ask how you can help 🤗

The table of projects which are currently supported.

If you'd like to add a new project to good-first-issue, you're more than welcome to submit a PR! There are a few components you'll need to submit:

• Update data/projects.json

  • Add your <project> as a property of projects in the correct alphabetical position with an object that includes a name, description, and a q (representing the GitHub search query).

• Update README.md by running npm run markdown

  • This will automatically update README.md with the new project's data.

You can pull your queries directly from a standard GitHub search! If you want to build something a bit more complex, you can use the advanced search tool if you want to build more specific custom queries: https://github.com/search/advanced

As a CLI, good-first-issue uses the Commander.js CLI framework. If you want to better understand how our CLI is built, commander.js is pretty well documented. Also used are Chalk for terminal coloring and boxen to simplify the output container implementation.

Good First Issue follows a relatively strict release process intended to ensure the spice flows.

We use both GitHub Labels and Milestones to track releases. Since project additions count as a minor release, we prefer to space those out and ship them individually rather than shipping many at once. This pace may be revised later, but for now, it introduces the need for a release queue and setting things up to be released ahead of them actually being released.

We use the release queue label and milestone to queue up PRs that have been reviewed and are ready to be released.

Once a PR is ready to be released, a milestone will be added that correlates to the SemVer version it will be released in. Ideally this will eventually be used for changelog tracking but for now it's just a good way to keep organized. To keep things tidy, once a new version has shipped the milestone will be closed out.

Prior to each release, whoever is releasing should be testing the release locally to ensure that the code is working as expected. This would include either running npm i -g or npm link in the PR branch and then testing whatever the PR is adding. Ensuring the experience isn't broken is vital.

It is worth noting that we limit the file we publish to npm with the files property in package.json. This property prevents code that's not explicitly listed from being shipped. We have had a situation where local testing and the published module differed because a PR was merged that added needed code in a directory that wasn't included. So, what works on your machine may not work for the end user.

To test locally, using the modules tests with npm test and trying out a few different commands (like the selector, a specific project, a failed project, and so on) is reccomended. For example:

npm i -g # This assumes your current working directory is the module's directory
good-first-issue # run the interactive CLI
good-first-issue react # test the react project
good-first-issue node # test the Node.js project
good-first-issue github # test the GitHub organization, `github`
good-first-issue github/semantic # test the GitHub repo, `github/semantic`
good-first-issue thisisntarealprojectorgithuborg

If you are interested in fixing issues and contributing directly to the code base, please see the document CONTRIBUTING.md.