react native
engineer
I build React Native infrastructure, multi-agent tooling, and LLM behavior controls that make the wrong output harder to produce.
WORK
Below the fold: current systems, active research, and the writing archive.
Selected projects
- Rorschach Protocol
Research initiative into LLM behavioral engineering — non-consciousness as design advantage.
- Theseus
Opencode agent harness — agents rewrite agents, skills replace skills, the loop tightens every run.
Elsewhere
Writing archive
Essays on LLM behavior, agent systems, and infrastructure decisions that survive contact with code.