One CLI. All models. Zero lock-in OpenClaude doesn't care if you use GPT, Kimi, MiniMax or Grok it just ships The model-agnostic, provider-free coding harness

wellp, thats cool.. time to add this to my macmini then :D

try it bro

will do and give my honest opinion <3

Can we add skills just like claude code

yep

wohoo

But, it still breaks too! With my local llm setup. I will retest it again tonight or tomorrow.

can you try in main branch? or kindly submit issue in repo please

I will retest it and keep you updated. So, far I liked it a lot and I'm expecting the multi-agakts setups soon. Keep up the good work guys!

I built a WebUI wrapper for OpenClaude, then Served up with tailscale to my iOS. I added agents, scheduled task, multi agent. It now crushes Claude code Appreciate all your work

Does have 'Computer Use'? Can I control it remotely, whether through a bot or some other method? 🤔

will release this feature soon

How safe are we using this?

so does openclaude maintain model consistency across different platforms or do devs need to manage compatibility nuances themselves?

we maintain it

When we have a plug-in to IDE like Visual Code or Intellij IDEA?

@hugocfranco it has alrady VS code plugin

feels like "lock-in free" just means you added an abstraction layer you maintain

DM intern I've got a deal for you

Nice—model-agnostic tooling is a big deal. Curious how you handle provider differences/costs across models; I’ve been using OpenStacks ( to sanity-check coding API plan comparisons before wiring things up.

Will this get complains from Anthropic by using Claude name:)

Can u connect opencode api ?

yes

I installed it and my bitdefender antivirus went crazy seeing everything as a threat 🫠🥲