A lightweight firewall system to classify and filter LLM prompts as safe or jailbreak using Ollama models. - View it on GitHub