To illustrate this, let’s take a simple script utilizing http-server as an example.
{
"scripts": {
"start": "http-server build",
}
}
Now, when trying to start the server on dynamic ports, you may assume you can just append the necessary argument -p 1234 to the command as follows.
npm start -p 1234
However, you’ll notice that nothing happens—respectively that the server will simply start on its default port, not the port you specified.
So, why is that?
The npm CLI itself also accepts arguments and parameters like -s in npm start -s for running in silent mode. However, a referenced/utilized script could also provide support for a -s argument. Thus the issue/problem npm is facing here is to separate and distinct whether a passed argument is intended for the npm CLI itself or the utilized script.
Solution
The npm CLI allows to pass additional arguments and parameters to utilized scripts by passing them after a double dash (--) as follows.
npm start -- -p 1234
npm now appends that content to the script before executing it, resulting in the following overall command.
http-server build -p 1234
Please be aware that npm will just append the arguments to the script and not pass it to each script you’re using in a chained command. So if you’re using multiple commands within a single script (e.g. combined via &&), the arguments will simply be appended to the end of your script, resulting in the arguments being used by the last command of that script only.
Differences between npm and Yarn
When using Yarn as your package manager, you can directly specify parameters without that extra double dash (--).
yarn start -p 1234
When you pass the arguments using a double dash nevertheless, Yarn even warns you and informs you about this fact.
From Yarn 1.0 onwards, scripts don't require "--" for options to be forwarded.
In a future version, any explicit "--" will be forwarded as-is to the scripts.
