The Bridge between Humans and AI Agents

Germany
Joined January 2026
have codex usage limits been decreased by openai?
76
computer agents retweeted
Helping founders get real feedback — today I reviewed Computer Agents with fresh eyes. Hope it helps! betabloom.app/reviews/WDb5FY…
1
1
2
96
computer agents retweeted
the model alone is no longer the product
681
527
307
7,783
1,284,633
Al agents shouldn't just chat.
2
68
just refactored our pricing page and added a new, powerful pricing calculator!
2
78
Vibe coding is fun until the task is bigger than one prompt: files change, deps break, tests fail, screenshots go missing, context gets lost. We’re building “Jira + cloud computers for vibe-coding agents” --> tickets, persistent workspaces, logs, tests, reviews, and human approval before shipping. Does that framing make sense, or is there a better way to explain it? computer-agents.com
1
68
you can create and run very high quality marketing campaigns with computer-agents.com
33
New onboarding finished!
32
working on a new onboarding ;)
1
1
43
next version of computer agents actually visually operates computers on the cloud.
2
68
complex diff views coming to computer agents ios and mac app!
1
69
next version of computer agents is going to be awesome!
1
3
67
theaustralian.com.au/busines… this is the most amazing story I've ever seen.
3
48
computer agents retweeted
What the frick, this is insane A Sydney tech founder used ChatGPT and genetic data (thanks to alphafold data) to design a personalized mRNA cancer vaccine for his dog Rosie after traditional treatments failed. Within weeks, a major tumor shrunk by about 50%, stunning scientists and giving Rosie a new lease on life!! The experiment shows how AI and mRNA technology could speed up personalized cancer treatments for both animals and humans. And you can do it yourself. Holy
45
65
15
770
123,492
working on much better context operations on the chat module today.
26
designing with maximum attention to every detail
41
it amazes me that meta is not able to produce better performing LLMs with all the capital they have...
META's internal "Avocado" model is disappointing, barely outperforming Gemini 2.5 in evaluations. The shocker: despite all the investments in LLMs, they're considering licensing and using competitor models like Gemini. Next level cash burn.
41
Very excited for Deepseek v4 release!
1
33