I race Hyrox. I wanted an app as beautiful as the sport deserves.
So I built Hyrox Tracker 🚀
Tired of asking friends "how'd you do?" on repeat?
Just grab their results in seconds. Save them. Have them offline forever.
Your crew, always with you.
https://t.co/SydoSczohe
#hyrox #iOS #fitness
In tour .claude system file settings place the following config https://t.co/f3F55wuJnX A friend of mine share with me and I think it's excellent to always have a sense of the tokens usages with your current workflow.
I see too much tech-industry content about how “AI will replace X role”—definitely for clickbait reasons. However, this may play a really dangerous role in our subconscious.
I see too many people overlooking the opportunity to shape their relationship with these tools and optimize their workflow to produce more value, more efficiently.
Today, we’re launching the @link wallet for agents. It lets you securely empower agents to spend on your behalf. Your payment credentials are never exposed and you approve every purchase.
https://t.co/TcvEiVNth9
This month has been crazy for sure. Going back into my journal I realized I spent a lot of time outside my office/home—trips and commitments. While I had a lot of fun it was definitely draining. Looking forward to mundane daily life.
I'm loving how descriptive and easy to reach are the usage cost from @Railway I've been enjoying this tool for a while now. I can easily see how much is costing my mistakes with memory management. 🤣
This is gold:
"When an agent implements an "authentication" bead, it focuses on making auth work correctly. Whether the login form has good visual hierarchy, whether the error messages are helpful, whether the flow feels smooth on mobile: these are orthogonal concerns requiring a different mode of attention. Trying to do both at once produces mediocre results on both."
I think this is true also for our cave-men mode of software development. 🤣
If the word "iteration" above wasn't explicit enough, let me tell you, the framework encourages you to iterate 13 or more times to check your beads, not counting the iterations on the plan. So yeah, that's a lot of time.
A few days ago, I stumbled upon this: https://t.co/icaGxj9o0y. It's another suggested framework like the one I shared before (compound engineering) to work with AI agents. It's super long; while I encourage you to read it, I will also go into some learnings I got from phases 1-5, which are intended for new projects.
With so much emphasis on iteration, it can be hard to know when to stop. The framework recommends switching to beads ONLY when the plan feels mostly stable and the remaining improvements are about execution structure.