🔥 Datasets and env wrappers for offline safe reinforcement learning

🚨 Generate and explore malicious QR codes for cybersecurity education and authorized penetration testing to enhance digital security awareness.

beta discord bot to automatically manage old school role based verification bots. proactive tracking, pinging, dming, reminders for those in purgatory and eventually regular auditing of those who should be and are not.

NeMo Guardrails is an open-source toolkit for easily adding programmable guardrails to LLM-based conversational systems.

A real-time notification system that monitors the Florida Highway Patrol incident database statewide. It sends alerts with location, remarks, and report time, logs all CAD calls, and provides follow-up updates showing previous incident details. The system standardizes location data and generates Google Maps links. Not affiliated with FHP.

Criminal Database manager

Code for Chain-of-Thought Hijacking

FAA Preliminary Aviation Accident and Incident Data from August 10, 2021 forward

Deploy once. Continuously improve your AI agents in production.

Real-time face recognition system to identify missing persons from CCTV streams using Flask, OpenCV, and InsightFace.

A Streamlit app for exploring content moderation with Llama Guard on Groq.

PyBullet CartPole and Quadrotor environments—with CasADi symbolic a priori dynamics—for learning-based control and RL

How good are LLMs at chemistry?

AI-Powered Crime Prediction & Safe Route Planner for Urban Safety. Empowering citizens with real-time AI insights for safer travel and smarter cities.

An open-source LLM engineering platform featuring observability, metrics, prompt management, a playground, guardrail and safety evaluation. Fully integrates with OpenTelemetry

JMo Security Suite - Terminal-first security audit toolkit with many tools, multi-target scanning, & compliance

An experimental system programming language guided by principles rather than precedent

Concept demonstration of Adversarial Prompt Vaccination (APV) — a self-healing framework for Large Language Models (LLMs) that uses adversarial prompt synthesis and micro-updates for continuous alignment and safety improvement.

A Discord bot that shares ban lists of scammers across multiple Discord servers

Reliability, Availability, Maintainability, Safety (RAMS) analysis program.