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

# 🛡️ 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`

Cookies

Voting for Moderator Bot

Moderator Bot