eve-forum-moderator/README.md

EVE Forum Moderator

AI-assisted EVE Online forum moderation tool - analyzes user posting history for rule violations.

Features

• Fetches user post history from the EVE Online forums (Discourse)
• Analyzes posts using Claude AI for potential rule violations
• Generates Discourse-compatible markdown reports
• Supports configurable time ranges and output options

Installation

Requires Python 3.11+ and uv.

# Clone the repository
git clone https://gitea.ext.ben.io/b3nw/eve-forum-moderator.git
cd eve-forum-moderator

# Install dependencies
uv sync

Configuration

1. Cookie Setup

Export your browser cookies after logging into forums.eveonline.com:

1. Log into https://forums.eveonline.com with your moderator account
2. Open Developer Tools (F12) → Application → Cookies → forums.eveonline.com
3. Copy the values for _forum_session and _t
4. Create a cookies.env file:

_forum_session=your_session_cookie_value
_t=your_auth_token_value

Alternatively, use a browser extension like "cookies.txt" to export in Netscape format.

2. LLM API Setup

Create a .env file with your LLM proxy API key:

LLM_PROXY_KEY=your_api_key_here

Usage

Review a User

# Basic usage - analyze last 30 days
uv run eve-mod review "SomePlayer"

# Specify time range
uv run eve-mod review "SomePlayer" --days 14

# Verbose output with detailed analysis
uv run eve-mod review "SomePlayer" --verbose

# Custom cookie file and output path
uv run eve-mod review "SomePlayer" --cookies ./my-cookies.env --output ./report.md

View Forum Rules

uv run eve-mod rules

Options

Option	Description
--days, -d	Number of days to analyze (default: 30)
--cookies, -c	Path to cookies file (default: cookies.env)
--output, -o	Custom output path for markdown report
--verbose, -v	Show detailed violation analysis
--max-posts	Maximum posts to fetch (default: 200)
--enrich/--no-enrich	Fetch full post content (default: enabled)

Output

Reports are generated in two formats:

1. Terminal - Rich-formatted summary with color-coded severity
2. Markdown - Discourse-compatible file saved to reports/ directory

Sample Report Format

## User Review: SomePlayer
**Period:** Last 30 days
**Posts Analyzed:** 47
**Posts with Violations:** 3

---

### Summary

| Date | Topic | Rule | Severity | Confidence |
|------|-------|------|----------|------------|
| 2026-01-03 | [Thread Title](url) | 1.2: Flaming | [MED] | 85% |

---

### Detailed Analysis

#### 1. [Thread Title](url)
**Date:** 2026-01-03 | **Category:** General Discussion

**[MED] Rule 1.2 - Flaming** (Confidence: 85%)

> "Relevant quote from the post"

*Explanation of why this violates the rule*

Forum Rules

The tool checks against the official EVE Forum Moderation Policy:

Prohibited Conduct (1.x)

• Trolling, Flaming, Ranting, Personal Attacks
• Harassment, Doxxing
• Racism & Discrimination, Hate Speech, Sexism
• Spamming, Bumping, Off-Topic Posting
• New Player Bashing, Impersonation, Advertising

Prohibited Content (2.x)

• Pornography, Profanity
• Real Money Trading (RMT)
• Discussion of Warnings/Bans/Moderation
• Private CCP Communications
• In-Game Bugs & Exploits
• Real World Religion/Politics

Other Rules (3-9)

• Non-Constructive Posting
• Abuse of Forum Tools
• Re-Opening Locked Topics
• Attacking CCP/ISD Staff

Development

# Run with verbose logging
uv run eve-mod review "TestUser" -v

# Check installed version
uv run eve-mod --version

License

Private - for authorized moderator use only.