How we think about
security.

VORO treats a codebase as a threat surface, not a finding list. Six dimensions — fund safety, access control, external risk, code integrity, dependency health, and agent autonomy — each scored by Bayesian confidence and weighted by proximity to real risk. The result is a map that shows where risk concentrates, not just where it exists.

The name comes from Georgy Voronoi, a Ukrainian mathematician whose tessellations partition space into regions of nearest influence. No gaps, no overlaps. The same geometry appears in crystal lattices, cell membranes, and the structure of the universe.

We apply his principle to security in an agentic world — where autonomous systems carry their own attack surfaces and the threat geometry shifts with every deployment. Partition the space. Score by proximity. Let the structure show what matters.