From Agentic Sessions to an Engine
How isolated coding agents, Git coordination, and separate post-processing became an execution engine for continuous security research.

Welcome to my corner of the web. I'm a software engineer and security researcher specializing in blockchain protocols. I work on Ethereum's Protocol Security team, expanding and automating AI-driven security research. Previously, I spent around three years at Stacks, advancing from hands-on property-based testing and fuzzing to leading the internal security research team. I also handled the bug bounty program, vulnerability triage, and security for the protocol and major releases.
How isolated coding agents, Git coordination, and separate post-processing became an execution engine for continuous security research.
Why security findings need a fresh context with the opposite mandate.
Why leader-centric agent swarms do not preserve distinct security-research techniques.
Dates show the latest post in each series.
A series about making agentic security research repeatable across roles, runs, and targets.