PariahBot reads its environment from hom.env (not .env) in the bot's directory. The setup scripts create it from .env.example and fill in the Discord and dashboard values, and the server log's message text, interactively:

DISCORD_TOKEN=your-bot-token
CLIENT_ID=your-application-id
GUILD_ID=
LOG_CHANNEL_ID=
LOG_CHANNEL_SUCCESS_ID=
LOG_CHANNEL_ERROR_ID=
TWITCH_CLIENT_ID=
TWITCH_CLIENT_SECRET=
YOUTUBE_API_KEY=
INSTAGRAM_ACCESS_TOKEN=
INSTAGRAM_BUSINESS_ACCOUNT_ID=
DASHBOARD_PORT=
DISCORD_CLIENT_SECRET=
DASHBOARD_HOST=
DASHBOARD_PUBLIC_URL=
SERVER_LOG_MESSAGE_CONTENT=

Keep this file private — it holds the bot token. It is gitignored, and should be mode 600 on a Linux host:

chmod 600 /opt/PariahBot/hom.env

The bot reads hom.env only at startup, so restart it after changing anything here, as in the table above.

The audit log is always written to logs/commands.log regardless; the LOG_CHANNEL_* values only add a Discord mirror. The bot needs View Channel and Send Messages in whichever channels you point them at.

This is the variable people get caught by, so it's worth being explicit:

While developing — set it to your test server's ID. Command changes appear the instant npm run deploy finishes, with no propagation delay.

In production across multiple servers — leave it empty. Commands register globally and appear in every server the bot joins.

Switching from guild-scoped to global has one trap: guild-scoped registrations do not disappear when you register globally, so the old test server ends up showing every command twice. Clear them first:

# from the bot directory, with GUILD_ID still set to the old server
node -e '
require("dotenv").config({ path: "hom.env" });
const { REST, Routes } = require("discord.js");
const rest = new REST().setToken(process.env.DISCORD_TOKEN);
rest.put(Routes.applicationGuildCommands(process.env.CLIENT_ID, process.env.GUILD_ID), { body: [] })
  .then(() => console.log("cleared guild commands"));
'

Then blank GUILD_ID and run npm run deploy.

Almost everything else. Alert destinations, voice hubs, role menus, auto-delete rules, mod/member/streamer roles, captcha, honeypot, giveaways, the streamer list and where its alerts post, the followed Instagram accounts and their channel — all of it is set per server, through slash commands, and stored in the database rather than in environment variables. See Server Configuration.

That's deliberate: the bot runs in multiple servers from one process and one database, so per-server settings can't live in a single global config file.