Matthias Lis
338 posts
Matthias Lis
@matthiaslis
building cool things with cool AI tools | 📍 Warsaw / Dubai
Warsaw Katılım Aralık 2015
202 Takip Edilen51 Takipçiler
Matthias Lis retweetledi
Matthias Lis retweetledi
@NousResearch @vikrantnyc I know I’m talking about my today experience
English
@matthiaslis @vikrantnyc That is a bug in the migrate tool which Vik did not use. Already tracked in a PR.
English
I just saw an AI murder another AI. Hermes has completely killed my OpenClaw on the same machine and made it unbootable. I think Hermes wants to be my sole assistant and not compete with OpenClaw. @NousResearch
English
@NousResearch @vikrantnyc u guys are a little bit tricky to propose renaming .openclaw dir to .openclaw-migrated “y” by default which obviously breaks openclaw agent 😅
English
@vikrantnyc Did you use the migrate tool or install Hermes fresh?
English
Ollama Cloud is moving fast! They've GLM 5.1, Gemma 4, MiniMax M2.7, Qwen 3.5, Kimi K2.5 etc all in one subscription.
Has anyone here subscribed yet? How's the experience so far?
ollama@ollama
MiniMax M2.7 is available on Ollama's cloud, and is licensed for commercial usage. Use it with OpenClaw: ollama launch openclaw --model minimax-m2.7:cloud Coding agents, such as Claude: ollama launch claude --model minimax-m2.7:cloud Chat with the model: ollama run minimax-m2.7:cloud
English
the only model running in a subscription model that is usable in either @NousResearch or @openclaw
Arena.ai@arena
GLM-5.1 by @Zai_org is now #3 in Code Arena - surpassing Gemini 3.1 and GPT-5.4, and now on par with Claude Sonnet 4.6. The first frontier level open model to break into the top 3. It’s a major +90 point jump over GLM-5, and +100 over Kimi K2.5 Thinking. Huge congrats to @Zai_org on pushing open model progress forward 🚀
English
@ollama, a quick question, when `ollama launch openclaw`
Okay, mid-writing this post, you just introduced it to your docs, crazy
English
AI either speeds up your workflow or wastes weeks of your time. You end up testing tools that spent more on marketing than the product. They show up first. Doesn't mean they're what you need.
English
@adamwathan I'm really looking forward, haven't got my invite yet, but waiting patiently
English
Quick ui.sh demo — generating multiple design ideas to choose from, no matter what tech stack you use:
English
@matthiaslis @robinebers @KainoaBHN Hi Matthias,
I have just the tool that you are looking for.
And it supports not only Codex, but Gemini, Claude etc as well. You can customise review criteria as well
Available as a GitHub app at the moment and standalone app is in development.
English
built all day with Opus 4.6
now had GPT 5.4 High Fast review it
guess what it said
English
@robinebers @KainoaBHN Yeah, so by "cool" I meant anything more structured than asking Codex to do a code review, so your skill fits it perfectly. Thanks.
English
@matthiaslis @KainoaBHN depends on what you mean by cool
I semi automate them with a skill I built: github.com/robinebers/ski…
English
GLM 5.1 has been my favourite model for OpenClaw.
It's capable and cheap... been enjoying it
Z.ai@Zai_org
Introducing GLM-5.1: The Next Level of Open Source - Top-Tier Performance: #1 in open source and #3 globally across SWE-Bench Pro, Terminal-Bench, and NL2Repo. - Built for Long-Horizon Tasks: Runs autonomously for 8 hours, refining strategies through thousands of iterations. Blog: z.ai/blog/glm-5.1 Weights: huggingface.co/zai-org/GLM-5.1 API: docs.z.ai/guides/llm/glm… Coding Plan: z.ai/subscribe Coming to chat.z.ai in the next few days.
English
@SieversJosua @Rasmic It’s much better, I tried gpt5.4 and it was stopping randomly in the middle of the task
English
@Rasmic Like more favorite than opus 4.6 back in the good old OAuth days?
I switched to gpt5.4 following the ban, but it is not working well at all for me..
English
@robinebers @KainoaBHN is there any cool way of doing PR reviews using codex app?
English
@KainoaBHN pretty much
1. it committed API keys into the repo (private luckily)
2. it hardcoded a webhook fallback key to hide a missing env var
3. it exposed a billing api without auth, making it public
absolutely fucking insane model
GPT 5.4 now cleaning everything up
English
English
I switched @openclaw again to GPT-5.4 and am DYING love this caveman software engineer personality upgrade 10/10 gonna keep her
English
the best @openclaw setup so far
ollama launch openclaw --model glm-5.1:cloud
English
I really thought all the posts about GPT 5.4 having no soul on @openclaw were fake, but I'm telling you it's even worse than that.
English