Niubi Guard
Open-source GitHub repository abuse detection and response system
...It focuses on spam, harassment, coordinated attacks, suspicious issue activity, and bot-like comment patterns. The project gives maintainers control over rules, allowlists, detection signals, confidence thresholds, and response modes. It includes both a command-line workflow and a web console for configuring scans and reviewing results. AI-assisted detection can be connected through OpenAI-compatible models, but review-only behavior is the default. The system is designed to help teams protect open-source projects without making enforcement policies opaque or automatic by default.