If I see one more T3 Code fork with a kanban for agents I'm gonna lose it

when Anthropic released their Functional Emotions paper, I gave it to Claude and asked for a song. tonight I asked Opus 5.5 to create a video for it. and it's breathtaking.

Introducing Claude Opus 5.5, the first model in our new Claude 5.5 family. It performs at the level of Claude Fable 5.1 for most tasks, and costs 40% less to run than Opus 5.

Put your photo in ChatGPT and say, Based only on this photo, put me on the tv show I look like I belong on and generate an image.

Omarchy is not going to adopt a formal Code of Conduct because those turned out to be struggle-session processes in the last turning. But we are going to play nice. We're going to be courteous, professional, and composed in our direct interactions with users, competitors, and even haters. Because that's what's going to pave the path to the win. That does not mean we have to act as vendors. Nobody is entitled to anyone else's free time or volunteer contributions. And we're not going to tolerate people who act like entitled customers buying a service contract from a company. Because we don't sell one, and we aren't one. This is going to be difficult sometimes. Because people are unreasonable! Especially on the internet! But if you return fire with fire, the only thing that'll burn is our reputation. The troll who triggered you into flashing your teeth won't face any consequences, but every spectator who saw it will judge you. I'm not going to tell you how to act anywhere else. You don't owe this organization a suit-and-tie demeanor at all times either. It's fine to be quirky, irreverent, silly, and even occasionally get upset. You do you. But when you're engaging with our fast-growing community, please keep these thoughts in mind. That's how we play nice but win. --- Posted this to the fast-growing Omarchy army of volunteers on our internal Basecamp.

There was a blog post that covered this exact reason: unix.foo/posts/it-will-never…

Finally got around to trying @t3dotcodes. Its quite nice. A few key things missing that will keep me on Codex Desktop (I'm not calling it ChatGPT) for now since I mostly use Codex anyway: - No message queue, only steer. I constantly use both in Codex. I think the Codex ui of being able to easily turn a queued message into a steer and reorder them is perfect. - No plan usage info. There is a really neat "usage" tab that shows you estimated API equivalent token cost, but thats not actually useful beyond novelty and contemplating the price reckoning that may come some day. I need to know how much I have left on my subs. - I must just be missing it... but no notification settings? I can get around this in a janky way with turn completion hooks for providers running on the same machine but this one is such an oversight that I am sure I am just not finding the settings. - No way to attach files/folders outside of the project directory to a message? @ lets me refer to files in the current project, but in Codex I can also click + to add files from outside or drag and drop them into the prompt field. Don't see a way to do either in t3, but again maybe just missing it. I can obviously just manually paste paths but thats tedious. What "attaching" means may vary by frontend but I assume most just stick the path in the prompt. - Vague uneasy feeling that something T3 Code ends up being different from Codex Desktop and I unknowingly gimp the model's abilities. I know they both use the same codex harness under the hood but unclear that EVERYTHING is the same. I assume for example computer use doesn't work, and its got its own slightly different set of plugins, introduces its own tools to integrate with the its own UI, etc. Like the context window IS different, though probably not in materially important ways. But there will always be the vague feeling of "what if this would have worked better Codex Desktop". Though I guess the other direction is just as valid a worry. Some of these already have active Github discussions and are being worked on, so I'm sure they will land soon enough. The last one I don't think is real but will be hard to shake non the less. If I wasn't waiting on a Codex reset right now I would probably have already just forked and patched some of these for myself haha. Anyway, awesome work so far @theo @jullerino! Definitely rooting for this project.

The new btca experience, now just a skill b/c it's way better: github.com/davis7dotsh/bette… Turns out the best way to let coding agents search github for context is to let them search github repos for context