Notifications
Clear all
Scoped and Ephemeral Credentials for Agents
1
Posts
1
Users
0
Reactions
12
Views
Topic starter
June 22, 2026 10:19 am
I've been reading a lot here about the risks of giving an AI agent broad, long-lived credentials. It makes sense, but I'm just starting with a local AI assistant in my home lab.
I want to set up a simple system where my agent can restart a failed service, but I know using my main SSH key or a root password is a bad idea. What's the most straightforward, minimal step I can take? I'm thinking about user permissions or a very limited API key, but I'm not sure where to begin practically.