Sheil

I’ve been spending a lot of time working with remote machines, and eventually got tired of jumping between SSH clients, terminals, SFTP tools, port forwarding, and AI assistants.

So I started building my own.

Sheil is an open-source, cross-platform tool for managing remote and device connections, with an embedded local AI engine for command generation and auto-completion.

The idea is pretty simple:

Remote infrastructure management + AI assistance, without making privacy an afterthought.

Sheil currently focuses on:

• SSH connections

• SFTP

• Port forwarding

• Remote/device management

• Local AI command generation

• AI command auto-completion

The local AI part is particularly important to me.

When you’re working with infrastructure, the context you’re giving an AI can be pretty sensitive. Hostnames, internal paths, commands, configuration, credentials accidentally pasted into a terminal, and other operational details aren’t things I necessarily want going to a third-party AI service.

With Sheil, the AI engine runs locally, so the goal is to make AI-assisted infrastructure work possible while keeping that context on your machine.

It’s still very early. There are currently 0 stars on GitHub, and that’s fine. I’m more interested in finding the first handful of people who actually use this and tell me where I’m getting it wrong.

I’m especially interested in feedback from:

• Developers managing multiple servers

• DevOps / infrastructure engineers

• Homelab users

• People managing remote devices

• Anyone who uses SSH heavily

If that sounds like you, I’d love to know:

What does your current remote management workflow look like, and what’s the most annoying part of it?

Website: https://sheil.hrdtr.dev
GitHub: https://github.com/Hrdtr/sheil

I’m building this in public, so I’ll share progress, technical decisions, and probably some mistakes along the way.