
一款轻量级、可配置的 Paper/Spigot 聊天审核插件:拦截不当言论(支持意大利语和英语),向管理人员发送包含完整消息的通知,并支持防绕过模式(普通 / 严格 / 极致)。
✨ 功能特性
? 核心功能
- 支持 Unicode 感知的黑名单匹配,适用于单词和短语(减少误报)。
- 管理员通知中包含完整的原始消息以及被检测到的匹配项 (
{message} + {match})。
- 可配置的防绕过模式:
normal — 仅基于模式的检测(最不激进)。
strict — 模式 + Leet 字符及重复字母规范化 + 仅插入式模糊检测。
ultra — strict 模式 + 去除元音检查(最激进)。
- 可自定义消息和管理员格式。
- 默认不自动禁言(插件专注于检测和通知)。
- 轻量级,兼容 Paper/Spigot 1.19+(已在 1.21.x 版本测试通过)。
?️ 防绕过技术
- Leet 字符规范化(例如
f4ck → fack)。
- 减少重复字母(例如
fuuuck → fuuck)。
- 滑动窗口仅插入式模糊检测(检测添加额外字母的绕过方式,同时避免将类似
duck 的词误判为替换绕过)。
- 去除元音检查(可选的
ultra 模式)。
❓ 为什么选择 bChatMod?
- 准确:通过单词/短语边界和 Unicode 感知模式,避免了许多误报。
- 可配置:为您的社区选择合适的激进程度。
- 透明:管理员在采取行动前会收到包含上下文的完整消息。
? 安装
- 构建或下载 JAR 文件,并将其放入服务器的
plugins/ 文件夹中。
- 启动服务器以生成默认的
config.yml。
- 编辑
plugins/bChatMod/config.yml 以自定义黑名单、消息和防绕过模式。
- 使用以下命令重新加载插件:
/bchatmod reload
⚙️ config.yml
# Anti-bypass mode: "normal" | "strict" | "ultra"
anti-bypass-mode: "normal"
# Debug (toggle true while tuning)
debug:
block-log: false
# Messages
player-block-message: "&7You cannot send that message. It contains blocked words."
staff-message-format: "&b[bChatMod] &7{player} tried cursing: &r{message} &7(matched: &b{match}&7)"
# Permission required to receive staff notifications
staff-notify-permission: "bchatmod.notify"
# Blacklist (example entries)
blacklist:
- "fuck"
- "shit"
- "example-offensive-term"
? 注意事项
短令牌(长度 ≤ 3)采用保守处理方式以避免误报(例如,“ass” 出现在 “basso” 中时不会被拦截)。添加短令牌时请谨慎。
在调试期间使用 debug.block-log: true 可以在控制台中查看哪些模式/候选词被匹配。
命令
命令
描述
/bchatmod reload
重新加载配置并重建黑名单模式
? 权限
bchatmod.notify — 接收管理员通知(默认:op)
bchatmod.reload — 重新加载配置(默认:op)
bchatmod.bypass — 绕过聊天过滤器(默认:op)
? 占位符
{player} — 玩家名称
{message} — 玩家尝试发送的完整原始消息
{match} — 触发拦截的子字符串或变体
&b[bChatMod] &7{player} tried cursing: &r{message} &7(matched: &b{match}&7)

A lightweight, configurable chat moderation plugin for Paper/Spigot: blocks profanity (Italian + English), notifies staff with the full message, and supports anti-bypass modes (normal / strict / ultra).
✨ Features
?Core Features
- Unicode-aware blacklist matching for words and phrases (reduces false positives).
- Staff notifications include the full original message and the detected match (
{message} + {match}).
- Configurable anti-bypass modes:
normal — pattern-based detection only (least aggressive).
strict — pattern + leet & repeat normalization + insertion-only fuzzy detection.
ultra — strict + vowel-stripped checks (most aggressive).
- Configurable messages and staff formatting.
- No automatic muting by default (plugin focuses on detection and notification).
- Lightweight, compatible with Paper/Spigot 1.19+ (tested up to 1.21.x).
?️ Anti-Bypass Techniques
- Leet normalization (
f4ck → fack).
- Reduce repeated letters (
fuuuck → fuuck).
- Sliding-window insertion-only fuzzy detection (detects extra-letter bypasses while avoiding substitutions like
duck).
- Vowel-stripped checks (optional
ultra mode).
❓ Why bChatMod?
- Accurate: word/phrase boundaries and Unicode-aware patterns avoid many false positives.
- Configurable: pick the level of aggressiveness for your community.
- Transparent: staff receive the full message for context before action.
? Installation
- Build or download the JAR and place it in your server
plugins/ folder.
- Start the server to generate the default
config.yml.
- Edit
plugins/bChatMod/config.yml to customize blacklist, messages and anti-bypass mode.
- Reload the plugin with:
/bchatmod reload
⚙️ config.yml
# Anti-bypass mode: "normal" | "strict" | "ultra"
anti-bypass-mode: "normal"
# Debug (toggle true while tuning)
debug:
block-log: false
# Messages
player-block-message: "&7You cannot send that message. It contains blocked words."
staff-message-format: "&b[bChatMod] &7{player} tried cursing: &r{message} &7(matched: &b{match}&7)"
# Permission required to receive staff notifications
staff-notify-permission: "bchatmod.notify"
# Blacklist (example entries)
blacklist:
- "fuck"
- "shit"
- "example-offensive-term"
? Notes
Short tokens (length ≤ 3) are handled conservatively to avoid false positives (e.g., "ass" inside "basso"). Use caution when adding short tokens.
Use debug.block-log: true during tuning to see which pattern/candidate matched in console.
Commands
Command
Description
/bchatmod reload
Reload configuration and rebuild blacklist patterns
? Permissions
bchatmod.notify — receive staff notifications (default: op)
bchatmod.reload — reload config (default: op)
bchatmod.bypass — bypass chat filter (default: op)
? Placeholders
{player} — player name
{message} — full original message the player attempted to send
{match} — the substring or variant that triggered the block
&b[bChatMod] &7{player} tried cursing: &r{message} &7(matched: &b{match}&7)
评论 0
登录后参与评分
还没有评论,来抢沙发吧