AI Moderator

[!(https://i.postimg.cc/x8G61ygD/minecraft-title.png)](https://postimg.cc/ctHRD3VX)

Stop fighting with 2012-style word filters. Start understanding what your players are actually saying.

[!(https://img.shields.io/badge/Platform-Paper-ffca28?style=for-the-badge&logo=minecraft&logoColor=white)](https://papermc.io) [!(https://img.shields.io/badge/Java-21-ed8b00?style=for-the-badge&logo=openjdk&logoColor=white)](https://www.oracle.com/java/) [!(https://img.shields.io/badge/Version-1. 0-blue?style=for-the-badge)](https://modrinth.com) [!(https://img.shields.io/badge/License-ARR-green?style=for-the-badge)](https://opensource.org/licenses/ARR)

## Key Features

• Actionable AI (Player Requests): Players can ask the AI for items, time changes, or commands, which generates an admin approval request (`` / ``).
• Technical AI: Admins can use `/aimod techai ` to let the AI dynamically change configs or execute console commands.
• Contextual Awareness: The AI reads the active plugins on your server and uses this knowledge to help players navigate custom menus or server features without revealing the raw plugin list.
• Multi-Model Support: Native integrations with Gemini, OpenAI, and Ollama.
• Customizable Heuristics: Advanced pre-filtering before passing text to LLMs.
• Punishment System: Configurable mute, kick, or ban actions per category.
• Review System: GUI and Chat-based review workflows for staff (`/aimod review`).
• Discord Integration: Webhook alerts for flagged messages.

"Wait, another chat filter?" 🤨 Nope. Traditional filters are dumb. They block "beach" because it looks like a bad word, but they miss a player being incredibly toxic without using a single swear word.

AI-Moderator is different. It uses a real brain (LLM) to understand context, intent, and tone. It knows the difference between a friendly joke and actual harassment.

## Why your staff will love this

### The "Triage" Logic (Fast & Smart) We don't waste your money. Every message first hits our lightning-fast local filter. Safe? Passes instantly. Obvious Spam? Blocked instantly. Suspicious or Toxic? Only then we ask the AI for a "second opinion".

### Conversation Memory The AI doesn't just look at one message. It remembers the last few things said in the conversation. It catches players who try to spread their toxicity across multiple messages to bypass filters.

### Automated Justice Configure your own rules. If a player reaches a certain "Violation Score", the plugin handles the rest: > 10 pts: Warning ⚠️ > 20 pts: 10m Mute 🔇 > 50 pts: Kick 👢 > 100 pts: Week-long Ban 🔨

## Features at a Glance

| Feature | Description | | :--- | :--- | | - Zero Lag | 100% Asynchronous. Your TPS stays at 20. 0, always. | | - AI Assistant | Players can use `/ai` to get answers to game questions. | | - Staff GUI | Review flags in a beautiful, intuitive in-game menu. | | - Discord Sync | Real-time alerts with embeds via DiscordSRV or Webhooks. | | - Toggleable Swears | Turn off swear moderation instantly in `filters.yml`. |

## Supported AI Engines Choose your favorite "brain": Google Gemini (Flash): The fastest and cheapest. Our top pick! OpenAI (GPT-4o-mini): The industry standard. Ollama: Run it on your own hardware for 100% privacy and $0 cost.

## Commands & Permissions

| Command | Action | Permission | | :--- | :--- | :--- | | `/ai ` | Ask the AI Assistant anything! | Everyone | | `/aimod review` | Open the Review Dashboard | `aimoderator.review` | | `/aimod stats` | Check AI performance | `aimoderator.stats` | | `/aimod reload` | Apply config changes | `aimoderator.admin` |

### Ready to clean up your chat? Download AI-Moderator now and let the machines do the boring work.

[ (https://github.com/4rubka/AI-Moderator) ] • [ (https://github.com/4rubka/AI-Moderator/issues) ] • [ (https://discord.gg/egg4X2yHVu) ]