Show HN: Fence – Jiminy Cricket for AI coding agents

Andrios, founder of hoop.dev (YC W21), announced Fence, an open-source tool that prevents AI coding agents from executing catastrophic commands like 'rm -rf ~' by reading intention rather than using a denylist. The tool, developed during a 20% side-project time, is currently used with Claude Code and Codex, with plans to support Cursor next.

Hi everyone, I'm Andrios, founder of hoop.dev YC W21 . 2 weeks ago, I told our eng team they could spend 20% of their week on side projects. The result: They came out with 3 open source projects. Tools we're using ourselves. Fence is the first one. It prevents an agent, confused or poorly prompted, to execute a catastrophic command before it happens. This could be 'rm -rf~’, ‘rm -fr~’ or even ‘sudo rm -rf $HOME’. It’s different from a denylist, because it reads the intention, not only the string. We use this for Claude Code and Codex. We want to see adoption before jumping to Cursor next. The team is very excited to release this, so please to drop any comments with feedback or issues you find: github.com/hoophq/fence Comments URL: https://news.ycombinator.com/item?id=48821039 https://news.ycombinator.com/item?id=48821039 Points: 1 Comments: 0