Amanda was not originally designed as an open-source project. Amanda's code is published for transparency, rather than to encourage people to run her themselves. You should use the official Amanda (https://amanda.moe/to/add) for day-to-day use as it is against the license to self host Amanda for public use. Nevertheless, the process to get a copy of Amanda up and running on your own machines is documented here, for those who want it.
Please read the entire file before actually doing anything. You may find that a later step, and as such the entire guide, is impossible for you to complete; Therefore, reading it first will save you time by not attempting the previous steps first.
Amanda depends on these services which are already hosted by other people. You'll need to sign up for them.
Create an application on Discord, then create a bot account with it. Note the token.
Sign up for the Chewey API and claim an API key.
Get a weeb.sh API token, somehow.
Amanda requires several components which must all be installed first. Not all of them need to be installed on the same machine.
Amanda stores all data in Postgres which includes arbitrary settings and other user related data and select data emitted by Discord's gateway.
Amanda uses Volcano (a Lavalink rewrite) as its audio processor. Download an appropriate version of Lavalink or any Lavalink compatible replacement such as Volcano, or build your own, and run it on a server. You should put it close to the Discord voice regions that you want to stream to.
You may use the provided application.yml file.
Amanda uses Yarn as a package manager instead of NPM, though to install it, you need to npm install -g yarn
Clone Amanda's repo.
Run yarn install.
With everything installed and running, you now have to configure it.
Open config.sample.js in the bot repo and fill in the details. You
must fill in everything. Rename the file to just config.js
Again, use the application.yml file from this repo.
Create a password for the Lavalink server, and put it in the file where directed.
Restart Lavalink to apply changes.
Ensure a user with the name amanda is created. The password can be whatever you chose earlier when setting up the bot config
You can import the Postgres schema you can find in the pgsql.dmp
It does contain some legacy tables that are no longer used at the moment that you can safely delete.