Writing / Technical field notes

What I learn while building.

Notes from projects where an implementation detail, measurement, failure, or connection changed how I understood the larger system.

Building ResearchComb to Keep Claims Tied to Sources

Why I built ResearchComb, how its workflows keep claims tied to sources inside Codex and Antigravity, and how to install the current v0.5.23 bundle.

Read article ↗

ResearchComb Open-Source Research Toolkit for Codex and Antigravity

I’m sharing ResearchComb — an open-source research toolkit designed for Codex and Antigravity, directly inside your AI chat. There’s no ResearchComb server, separate app, or API key to configure.

Read on LinkedIn ↗

What The Imitation Game Taught Me About System Design

How a repeated phrase turns an impossible search into a constrained problem—and what that teaches us about invariants, protocols, observability, debugging, and security.

Read article ↗

Connecting My Maze Visualizer to Redis

How two separate C++ projects became one useful system through serialization, POSIX sockets, RESP, validation, and persistent storage.

Read article ↗

Why Dijkstra Beat A* in My Tiny Maze

A practical look at why the theoretically smarter algorithm can lose on a small grid—and what profiling taught me about heuristics, overhead, and hardware reality.

Read article ↗