16 y/o at NCSSM. I work on agentic harness engineering, reasoning in latent space, recursive self-improvement, and synthetic data.
wizard -- Self-extending autonomous coding agent in one Rust binary. Any provider (OpenAI-compatible, Anthropic, xAI) or fully local via llama.cpp. Live /evolve self-modification, MCP, messaging gateway, built-in bench. Ran its own AHE evolution loop under Grok 4.5 to drive the harness to a 100% pass rate.
agentic-harness-engineering (contributor) -- Observability-driven automatic evolution of coding-agent harnesses. NexAU-AHE reaches 84.7% ± 2.1 pass@1 on Terminal-Bench 2, beats Codex/ACE/Training-Free GRPO.
reverie -- JAX + Rust successor to Coconut. Depth-supervised halting calibrates latent compute to problem depth -- reasoning steps match hop count exactly, with a causal γ-ablation.
FUSION -- Multi-agent LLM debate framework in Rust. Now a workspace crate powering wizard's /fusion command.
candor-bench -- Honesty, sycophancy, calibration, and factuality evals for LLMs.
spore -- The most minimal Rust coding agent: one tool (bash), spawns and rewrites itself.
Python, Rust, PyTorch, JAX, SGLang, Nix, some TypeScript, some C++, some Lua. NixOS daily driver.




