🛡️ Moderator Bot

Free AI-powered moderation for Discord. Detects nudity, gore, scams, and other violations in messages, media, and avatars. Fully configurable via slash commands.

📌 Features Overview

NSFW Filtering Detects explicit or harmful content in uploaded media (images, GIFs, Lottie, APNG), Tenor GIFs, and user avatars. Categories include nudity, sexual content, graphic violence, gore, and self-harm-related material.
Strikes Escalating punishment system that applies timeouts, bans, or other actions based on user strike count. Fully configurable with custom durations, action cycling, and optional DM notifications.
AI Moderation Uses OpenAI models to automatically moderate messages based on server rules. Supports autonomous moderation, contextual analysis, batch scanning, and rule-based enforcement. By default, AI moderation runs in report mode and is triggered when users @mention the bot.
Scam Detection Detects scam and phishing messages using pattern matching, Google Safe Browsing, PhishTank, link unshortening.
Banned Words Blocks slurs or custom-defined banned words. Supports layered punishment and integrates with the strike system.
Logging Tracks major server events including joins, leaves, bans, deletions, edits, timeouts, and invite usage.
Custom Settings All moderation behavior is configurable through slash commands. Includes rule definitions, exclusion settings, thresholds, model selection, and more.
Private API Pool Allows users to contribute OpenAI API keys for increased capacity. Keys are encrypted and only used for moderation purposes.

🔺 Strike System

Automatically escalates punishments:

1st Strike → 1d timeout
2nd Strike → 7d timeout
3rd Strike → Ban

🧠 Fully customizable:

/strikes add_action for custom escalations
/settings strike-expiry to auto-expire strikes
cycle-strike-actions to reuse actions when needed
dm-on-strike to DM users when a strike is issued
Logs go to strike-channel

🤖 AI Moderation (Batch & Autonomous)

Uses OpenAI to moderate user behavior based on configured rules:

Run on interval (aimod-check-interval) or mention (aimod-mode)
Default mode is report: users @mention the bot to trigger AI moderation
Supports autonomous moderation (autonomous-mod)
Context-aware analysis (contextual-ai)
Custom rules via /ai_mod rules_set
Custom actions: aimod-detection-action

🔧 Settings:

rules: Defines the rule set used by AI
aimod-model: Choose OpenAI model (e.g. gpt-4o)
autonomous-mod: Toggle for autonomous mode

🛠 Commands:

/ai_mod toggle, /ai_mod add_action, /ai_mod view_actions

🖼 NSFW Filtering

Detects NSFW in:

Uploaded media (images, GIFs, stickers)
Tenor GIFs (check-tenor-gifs)
Profile pictures (check-pfp)
Lottie animations and APNGs

🚨 Categories:

nsfw-detection-categories: e.g., sexual, violence_graphic
threshold: Confidence threshold (lower = stricter)

🔨 Actions:

nsfw-detection-action: Strike, kick, delete, etc.
nsfw-pfp-action: Kick, timeout, etc.
nsfw-pfp-message: Message shown to flagged users
unmute-on-safe-pfp: Auto-unmute after avatar change

🔧 Commands:

/nsfw add_action, /nsfw add_category, /nsfw view_actions, /nsfw set_threshold, /nsfw view_threshold

❌ Scam & Link Protection

Protect against scams and phishing:

Match messages and URLs manually
Link checks via Google Safe Browsing and PhishTank
Unshortens and verifies destination
Actions: scam-detection-action, delete-scam-messages

🔧 Toggles:

check-links: Enable link checks
exclude-scam-channels: Channels to ignore

🛠 Commands:

/scam settings, /scam view, /scam list_patterns, /scam list_urls

💬 Banned Words System

Uses built-in slur list (use-default-banned-words) or custom words
Actions: banned-words-action (delete, strike, etc.)
Layered enforcement and logging in monitor-channel

🛠 Commands:

/bannedwords add, /bannedwords remove, /bannedwords add_action, /bannedwords view_actions, /bannedwords clear

📊 Monitoring & Logging

Logs: Joins, leaves, edits, deletions, bans, kicks, timeouts
Tracks deleted message authors even if uncached
Invite tracking for new members
Output channel: monitor-channel

Command: /monitor set

⚙️ Settings Snapshot

Name	Type	Description
`strike-channel`	TextChannel	Logs strikes
`nsfw-channel`	TextChannel	Logs NSFW previews
`monitor-channel`	TextChannel	Logs general events
`api-key`	str (encrypted)	OpenAI key for AI/NSFW moderation
`strike-expiry`	TimeString	Duration before strikes expire
`cycle-strike-actions`	bool	Loop fallback strike actions
`dm-on-strike`	bool	DM users when they receive a strike
`strike-actions`	dict	Action mapping per strike level
`check-pfp`	bool	Scan avatars for NSFW
`nsfw-detection-categories`	list[str]	Which NSFW categories to detect
`threshold`	float	Detection sensitivity
`nsfw-pfp-action`	list[str]	Action on NSFW avatars
`nsfw-pfp-message`	str	Message on NSFW avatar detection
`unmute-on-safe-pfp`	bool	Auto-unmute on safe avatar change
`check-tenor-gifs`	bool	Scan Tenor GIFs for NSFW
`use-default-banned-words`	bool	Use built-in slur list
`banned-words-action`	list[str]	Action on banned words
`exclude-channels`	list[TextChannel]	Channels excluded from checks
`delete-scam-messages`	bool	Auto-delete scam messages
`scam-detection-action`	list[str]	Actions for scam messages
`check-links`	bool	Enable URL safety checks
`exclude-scam-channels`	list[TextChannel]	Skip scam checks in these channels
`rules`	str	Server rules used by AI moderation
`aimod-detection-action`	list[str]	Action when AI flags content
`autonomous-mod`	bool	Enable autonomous AI moderation
`aimod-model`	str	Model used for AI mod
`aimod-check-interval`	TimeString	How often to run AI moderation
`contextual-ai`	bool	Enable context-aware moderation
`aimod-mode`	str	`report` or `interval` mode
`no-forward-from-role`	list[Role]	Roles that can’t forward messages

💬 Command Index

General

/help, /settings, /api_pool, /monitor

Strikes

/strikes get, /strikes remove, /strikes clear
/strikes add_action, /strikes remove_action, /strikes view_actions

NSFW

/nsfw add_action, /nsfw remove_action, /nsfw add_category, /nsfw view_actions
/nsfw set_threshold, /nsfw view_threshold

AI Mod

/ai_mod toggle, /ai_mod rules_set, /ai_mod view_actions, /ai_mod add_action, /ai_mod remove_action

Banned Words

/bannedwords add, /bannedwords remove, /bannedwords defaults, /bannedwords add_action, /bannedwords view_actions

Scam

/scam settings, /scam check_links, /scam add_message, /scam add_url, /scam list_patterns, /scam list_urls