# About VineScale · Founded by Gary Mathis

> VineScale is a safety gate for AI coding agents, founded by Gary Mathis and now in free beta. Why it exists, where it's going, and who writes the blog.

Web page: https://vinescale.com/about/

[Home](https://vinescale.com/)

## About VineScale

VineScale is a safety gate for AI coding agents such as Claude Code, Codex and Cursor. It knows a team's production databases and infrastructure, and warns about or blocks an agent's action based on what that action would break. VineScale was founded by **Gary Mathis**, and its gate for Claude Code is in free beta, with beta support for Codex, Cursor, Antigravity and Grok Build.

### Why VineScale exists

Coding agents run real commands with whatever access a developer's terminal has. Public reports describe agents that dropped production databases, ran `terraform destroy` on live infrastructure and deleted repositories, sometimes after being told not to. Our [roundup of AI coding agent incidents](https://vinescale.com/blog/ai-coding-agent-incidents/) collects the cases that have been reported publicly.

Today's safeguards, like permission rules, hooks and sandboxes, judge a command by its text or by where it writes. They can't tell a harmless table from one that another service reads. VineScale closes that gap by judging an action by what it would actually touch.

### Where it's going

- A free gate for Claude Code, Codex, Cursor, Antigravity and Grok Build first, for TypeScript, Python and Go projects. It's in beta now.
- Then a paid Team tier, with shared rules for the whole team and an audit log.

[Try the free beta](https://vinescale.com/beta/), and [join the waitlist](https://vinescale.com/#waitlist) for early access to the Team tier.

### Who writes the blog

Gary Mathis, VineScale's founder, is the author of the articles on the [VineScale blog](https://vinescale.com/blog/). They are practical guides on running AI coding agents safely around production systems, with sources linked for every factual claim.