╔═══════════════════════════════════════════╗
║                                           ║
║   ██╗  ██╗ ██████╗   ██████╗ ██╗          ║
║   ██║ ██╔╝ ██╔══██╗ ██╔═══██╗ ██║         ║
║   █████╔╝  ██████╔╝ ████████║ ██║         ║
║   ██╔██╗   ██╔══██╗ ██╔═══██║ ██║         ║
║   ██║ ██╗  ██║  ██║ ██║   ██║ ████████╗   ║
║   ╚═╝  ╚═╝ ╚═╝  ╚═╝ ╚═╝   ╚═╝ ╚═══════╝   ║
║                                           ║
╚═══════════════════════════════════════════╝

---

About Me

Hi, I'm Cao Hanzhe (Kral) — a CS student, AI researcher, and open-source enthusiast.

• 🔭 Building Multi-Agent Systems that reason, debate, and collaborate to solve complex real-world problems
• 🧠 Advancing Reinforcement Learning — from RLHF reward modeling to agentic RL with multi-turn reasoning
• 🤖 Bridging the gap between simulation and real-world robotics — sim-to-real transfer, embodied AI
• 🌐 Pushing the frontier of LLM Reasoning — test-time compute scaling, search-augmented generation, tool-use agents
• 🏗️ Creator of MingJian (明鉴) — an evidence-driven multi-agent simulation platform for strategic decision-making
• 📬 Reach me at: cajd6876@gmail.com

What I Do

🤖 Multi-Agent Orchestration Designing agent systems where multiple LLMs collaborate through debate protocols, evidence chains, and structured reasoning — not just simple tool-calling.	🧠 Reinforcement Learning Implementing and fixing core RL algorithms — from PettingZoo parallel environments to LinUCB contextual bandits. Contributing fixes upstream to pytorch/rl and Pearl.
🦾 Robotics & Sim-to-Real Working with robosuite and NVIDIA IsaacLab to build robust simulation pipelines that transfer to real robots. Fixing core physics engine bugs and resource management.	🛡️ AI Safety & Evaluation Building automated safety checks — prompt injection detection, red-teaming frameworks, and LLM evaluation harnesses. Contributing to Giskard AI safety platform.

Featured Project

🏗️ MingJian (明鉴) — Multi-Agent Decision Platform

AI-powered multi-agent platform for evidence-driven scenario simulation and strategic decision-making

• ⭐ 19 stars · Python · FastAPI
• 🏛️ Supports corporate and military strategic domains
• 🎭 Multi-agent debate protocol with evidence chains
• 📊 Real-time scenario simulation engine
• 🔗 github.com/dashitongzhi/MingJian

Tech Stack

Languages

AI / ML

Infrastructure

GitHub Stats

Open Source Contributions

Active contributor to 30+ AI and agent projects across GitHub — fixing core bugs, adding safety features, and improving developer experience

Category	Projects	Highlights
🤖 Agent Frameworks	rllm, notte, Composio/agent-orchestrator, stakpak/agent	Core bug fixes, session management, async improvements
🧠 Reinforcement Learning	pytorch/rl, facebookresearch/Pearl, alibaba/ROLL	Fixing PettingZoo parallel env bugs, LinUCB tensor squeeze, agentic LR scheduler
🦾 Robotics	robosuite, IsaacLab, SmolVM	Resource leak fixes, docstring corrections, sim-to-real improvements
🔧 AI Infrastructure	Kokoro-FastAPI, any-llm, Art, burr	Error message fixes, kwargs passthrough, install automation
🛡️ AI Safety	Giskard-AI, Agent-R1	LLM-based prompt injection detection, red-teaming checks
📦 Dev Tools	visidata, cc-switch, hermecore, go-micro	Shell command fixes, metadata parsing, CI improvements

Achievements

• 🏆 Starstruck — Repository earned 16+ stars
• 🦈 Pull Shark — Merged 30+ pull requests across major open-source projects
• 📊 330+ contributions in the last year
• 🌍 Contributed to projects from Meta, PyTorch, Alibaba, NVIDIA, Apache and more
• 🚀 Built and maintained MingJian — a production-grade multi-agent platform

Contribution Graph

Ask Me About

Currently Working On

• 🏗️ MingJian v2 — Enhanced multi-agent debate protocol with evidence chain validation
• 🧠 Agentic RL — Multi-turn reinforcement learning for LLM agents
• 🦾 IsaacLab Contributions — Improving sim-to-real transfer pipelines
• 🛡️ Prompt Injection Detection — Building automated LLM safety evaluation tools

Quote

"The question of whether machines can think is about as relevant as the question of whether submarines can swim." — Edsger W. Dijkstra

Let's Connect

I'm always open to collaboration on multi-agent systems, RL research, and robotics projects.

If you're building something in the AI agent space — let's talk! 🚀