Jev Moderation Bot

# Jev Moderation Bot

A Discord moderation bot built with Python and TypeSafe AI (Jev System One). It filters spam and scam links in real time, escalates offenses automatically, and lets moderators profile members based on their message history.

## What it does

- **Real-time link and spam moderation**: Analyzes messages as they arrive. If a message is flagged as spam or a scam link, it gets deleted immediately with a reason logged to Discord's server audit log.
- **Progressive discipline**:
  - 1st offense: Warning DM
  - 2nd offense: Final warning DM
  - 3rd offense: 10-minute timeout (customizable)
  - 4th+ offenses: 1-hour timeout (customizable)
- **User profiling (`/profile`)**: Analyzes a user's recent messages (10 to 100) and scores them across several metrics: scam risk, spamminess, noobness, toxicity, and helpfulness. Assigns a persona (e.g. beginner, regular, spammer, troll) and suggests next steps.
- **Rolling message cache & channel fallback**:
  - Passively logs incoming chat messages into a local SQLite database with timestamps.
  - If a member is new or has no cached history, moderators can pick a channel from a dropdown to fetch recent messages on the fly. All scanned messages are cached so subsequent lookups are instant.
- **Mod logs with 1-click actions**: Posts alerts to a designated `#mod-log` channel with buttons to pardon false positives or ban the user.
- **Dynamic false-flag learning**: When an admin pardons a message, that message is saved as a safe precedent and included in future AI checks for that server, preventing repeated false alarms.

## Commands

| Command | Arguments | Permission | Description |
| :--- | :--- | :--- | :--- |
| `/profile` | `user`, `[message_count]`, `[channel]` | Moderate Members | Builds an AI profile for a member sc