Selected Work

Open-source tools, AI security research, publications, and talks.

A curated index of public-safe work across developer-first security tooling, AI security research, AWS and CSA publications, and conference sessions.

Open source

Automated Security Helper

Developer-first security scanning for local and CI/CD workflows. Current headline metric: 200k+ monthly clones.

AI security research

AI Leak Watch

A public dashboard that tracks potentially exposed AI provider keys on GitHub and explains why secret leakage gets more serious in agentic systems.

AI agent evaluation

SIR-Bench

A benchmark for evaluating whether security incident response agents conduct real investigation or simply repeat alert context.

Publications & Research

Selected public writing, research, and framework contributions.

Talks

Selected conference sessions, workshops, videos, and demos.

2026 / Upcoming

RBLN East: SIR-Bench

Evaluating investigation depth in security incident response agents, with Cristian Leo.