Gale Fagan

I build AI agents for production and research organizational trust and AI at work.

I’m Principal AI Architect at Brightflag, a Wolters Kluwer company, responsible for AI adoption across the company. My work combines hands-on architecture with forward-deployed engineering. I’m also a doctoral candidate at Edgewood University, studying what people do when their employer’s sanctioned AI tools can’t do the job, and why they keep quiet about it. My security research examines malware campaigns in the AI agent ecosystem.

Before this I was CTO at dv01, where I led more than 70 engineers and delivered $3 million in cloud savings. I’ve helped scale companies past $100M in revenue, led the Nomad engineering team at HashiCorp, and spent six years at Apple making the factory lines that build iPhones run faster.

Lately

• October 2026. I presented “Librarian Catches Thief” at SCORED ’26 in Prague, co-located with OpenSSF Community Day Europe. Malware campaigns in AI agent skill registries copy their own files hundreds of times, and plain document similarity finds the copies. The paper is free to read, and the slides are on the session page.
• September 2026. plugin-place is public: my own Claude Code plugin marketplace, with every vendored skill keeping its author and license.
• 2026. Plugin Librarian, the tool behind the paper, is open source. It tells you which of your Claude Code plugins overlap before you install one more.