I Built 174 AI Agents That Fight Each Other.

This is a submission for the DEV April Fools Challenge What I Built Most multi-agent systems make agents cooperate. I made mine fight. Meet BlackSwanX — an adversarial intelligence engine where 200 citizen AI agents argue, panic, and emotionally spiral while a BlackSwan Assassin tries to murder the consensus. It runs 100% locally on Ollama. Zero API cost. Maximum chaos. I deployed a Vedic Astrologer, a Panic Seller, a Chaos Mathematician, a Gen Z Culture Decoder, and a Street Smart Hustler (who will tell you "your pitch deck is pretty, show me your bank account") to predict the future. Together. By fighting. This solves zero real-world problems elegantly. It just finds where the crowd is wrong. Demo 👉 GitHub Repo — BlackSwanX Quick start (2 minutes): git clone https://github.com/Kalki-M/BlackSwanX.git cd BlackSwanX ollama pull llama3.2:3b && ollama pull phi4:14b pip install -r requirements.txt bash start.sh Example run — "Will NVIDIA crash when the AI bubble pops?": Kill Shot: