There’s been a ton of hype around AI and developer workflows, but when it comes to Kubernetes, not much has really happened yet - and for good reas...
For further actions, you may consider blocking this person and/or reporting abuse
been cool seeing steady progress on this. feels like the guardrails are what makes the difference for me - otherwise its chaos. interested how much smoother itll make my day-to-day once i give it a try.
Thanks! We always wanted to give DevOps a way to provide guardrails for developers, now we just expanded it to AI agents as well 😄
Any feedback is very welcome 🙌
Really awesome!
😊
This is a really fun one. Bookmarking it to give it a spin later on.
Yesss, just today I was looking into Cyclops MCP. These are great
🧡
Thanks for sharing!
😁
Cool stuff
🙌
Very cool feature. Keep it up!
🙌
Sensational! Finally, a way to use LLMs with Kubernetes without turning everything into a mess.
😁😁
honestly love seeing moves like this but gotta say k8s always seems like a beast - you think these guardrails actually make ai agents less risky
The core idea is this: instead of letting an AI write raw YAML or poke at your cluster directly (which is risky), Cyclops gives DevOps teams a way to define safe, reusable, and configurable templates through our own custom Kubernetes resource. These custom resources become the only thing the AI can see or interact with.
So yes, the "guardrails" are real. They limit what an AI can do to what's already been approved by humans, using good patterns. That massively reduces the risk of misconfiguration or breaking things in production.
This is actually the same concept we have been building for developers for some time now, now just expanded to include AI agents!
👀 👀
😄
Good stuff
🙌
Super cool!
Thanks!
Amazing
🫶
Vibe reading
😎