Veto is runtime authorization for AI agents. Before an agent runs a tool call, Veto checks it against the policies its developer set: which tools it may call, with what parameters, how often, and when. Nothing is allowed by default. Every decision is made at the edge in milliseconds and written to an append-only audit log.

I co-founded Veto in spring 2026 with Michael Malyshev and write all of its software: the policy engine on Cloudflare Workers, the REST API, the Node.js and Python SDKs on npm and PyPI, the middleware for Model Context Protocol (MCP) servers, the dashboard and the docs. YC S26 application: top 10%.