Sandboxing ruby projects from external changes is imho the sensible choice.

This is true, of course, for any development environment. It’s quite frustrating to come back to a project after some time and find it broken because it depended on a particular system configuration that just worked at the time.

One of the underlying remedies for ruby is to use a ruby version manager like rbenv. So on macOS the following will do the trick.

Step 1. Install rbenv with ruby-build

brew install rbenv

Step 2. Initialise your shell with rbenv

I use bash-it, thus

bash-it enable plugin rbenv

But, it is otherwise achieved via rbenv:

# Run this and follow the instructions to set up rbenv integration with your shell
rbenv init

Step 3. Verify rbenv is setup

# assume we've opened a new terminal
curl -fsSL https://github.com/rbenv/rbenv-installer/raw/master/bin/rbenv-doctor | bash

Step 4. Install recent ruby

Now install a ruby version via rbenv.

# list the most recent stable version:
rbenv install -l | grep -E '^\s*[0-9\.]+$' | tail -n 1

# install that ruby version globally. e.g.,
rbenv install 2.5.1
rbenv global 2.5.1

Step 5. Project initialisation

# init project dir and cd into it
mkdir my-new-project
cd $_

# init .ruby-version
rbenv local 2.5.1

An alternative / complementary solution, I use direnv. See direnv’s ruby wiki page for further tips.