[empath75]

I just want to point out _how easy it is_ to build any of these tools yourself.

I got inspired by nano-claw and built on some of it's ideas to build a whole k8s hosted autonomous agent platform and got it into production in 2 weeks. It's just some api calls and container orchestration. The only hard problem _and it is hard_, is securing it, because you basically have to treat the agents as potentially malicious.

[dsjoerg]

It's because of this hard problem that I'm thinking I should keep using Viktor (getviktor.com) instead of running this OpenViktor.

The company may not do a perfect job of security either, but I figure they'll do a better job than I can as a solo practitioner.

[openviktorpls]

u mentioned weighing OpenViktor against staying on Viktor for the security tradeoffs - did you happen to clone the repo before it went down? I'm trying to get a copy to evaluate it myself. would really appreciate it if you could share - [e-mail]

[fc417fc802]

I doubt it makes a difference. The primary risk is the agent exfiltrating your private data. That's going to exist either way.

Essentially anything you give it access to should be considered inside the same security boundary. Which is quite unfortunate if you want it to respond to emails for you and also query the internet at large.

[zggf]

Agree, anything that agent has access to is like giving it to malicious user. Especially when agent is exposed to different users that should have different permission levels