Content Filter

Courier scans mail for prohibited content before it’s sent — hate speech, slurs, and anything else you want blocked. It’s enabled by default with a bundled word list, and you can add your own patterns.

How It Works

When a player sends mail, Courier checks the subject, body, and any attached item names, lore, and book contents against a list of regex patterns. If something matches, the mail is blocked and the player sees an error.

Admins with the bypass permission skip the filter entirely.

Configuration

The filter lives in your main config file as the filter section.

File: config/courier/config.conf

filter {
  enabled = true
  action = BLOCK
  useBundledList = true
  customPatterns = []
  bypassPermission = "courier.filter.bypass"
  checkSubject = true
  checkBody = true
  checkItemNames = true
  checkItemLore = true
  checkBookContents = true
}

Setting	Default	What It Does
`enabled`	`true`	Turn the filter on or off
`action`	`BLOCK`	What to do when content matches: `BLOCK` rejects the mail, `CENSOR` replaces matched text with asterisks
`useBundledList`	`true`	Include Courier’s built-in hate speech patterns
`customPatterns`	`[]`	Your own regex patterns to match against
`bypassPermission`	`"courier.filter.bypass"`	Permission node that skips the filter
`checkSubject`	`true`	Scan mail subject lines
`checkBody`	`true`	Scan mail body text
`checkItemNames`	`true`	Scan custom names on attached items
`checkItemLore`	`true`	Scan lore text on attached items
`checkBookContents`	`true`	Scan the contents of attached written books

Custom Word List

Add your own blocked words and patterns in a separate file:

File: config/courier/filter_words.txt

# Custom filter patterns for your server
# One pattern per line. Supports regex.
# Lines starting with # are comments.

badword
anotherbadword
custom[s$5]lur

This file is auto-created on first launch with a template and instructions. Patterns here are combined with the bundled list (if enabled).

Pattern Format

Patterns use Java regex syntax. Courier automatically adds word boundaries (\b) around your patterns so they don’t match inside other words.

Simple patterns match exact words:

badword

Character classes catch leet-speak variants:

b[a@4]dw[o0]rd

Common substitution classes:

[i1!|l] — matches i, 1, !, |, l
[e3] — matches e, 3
[a@4] — matches a, @, 4
[o0] — matches o, 0
[s$5] — matches s, $, 5

All matching is case-insensitive.

Filter Actions

BLOCK (default) : The mail is rejected entirely. The player sees an error message and nothing is sent.

CENSOR : Matched words are replaced with asterisks. The mail is still sent, but the prohibited content is hidden.

Bypass Permission

Players with the courier.filter.bypass permission skip the filter. By default, ops (level 4) bypass automatically.

You can change the permission node in config:

filter {
  bypassPermission = "courier.filter.bypass"
}

Or set up a LuckPerms group:

/lp group admin permission set courier.filter.bypass true

What Gets Scanned

The filter checks five parts of each mail, each controlled by its own toggle:

Subject line (checkSubject) — the mail title
Body text (checkBody) — the main message content
Item names (checkItemNames) — custom display names on attached items
Item lore (checkItemLore) — tooltip text on attached items
Book contents (checkBookContents) — pages of any attached written books

If any check finds a match, the whole mail is blocked (or censored, depending on action).

Error Messages

When mail is filtered, the player sees:

Primary: “Your message was blocked by the content filter.”
Detail: “Prohibited content detected in {location}.”

Where {location} is one of: “subject”, “message body”, “item name”, “item lore”, or “book contents”.

These messages are customizable in messages.conf under the contentFiltered and contentFilteredDetail keys.

Reloading

Filter patterns reload when you run:

/mailadmin reload

This picks up changes to both config.conf (filter settings) and filter_words.txt (custom patterns).

Next Steps

Moderation & Safety — rate limiting, blocking, and other abuse prevention
Main Config — all config.conf settings
Permissions — full permission reference