Manupa

My blog on shadcn/ui developed by the brilliant design engineer @shadcn is finally here. A technical analysis of tools, techniques and best practices that made it the most loved component library of 2023. manupa.dev/blog/anatomy-o… #buildinpublic #nextjs #React

@kevin2kelly @dwarkesh_sp Where can I find this ? Your books changed my life. Especially your writings on cybernetics and swarm intelligence got me into AI and CS in 2021

@manupadev @dwarkesh_sp HA! But thanks. I've already been on Dwarkesh's podcast, when he was starting out. He was already asking hard questions.

Who should I interview on my podcast? Open to more AI, but also to random history/econ/etc professors that I might not have heard of before.

Skills should be: - Concise - Responsible for one thing, not multi-step - Composable - Progressively disclosed - Harness-agnostic What else? Or - what did I get wrong?

Let’s go🚀

I am absolutely more productive using agents. I don't know the factor but it's large. However much of that productivity is spent tuning the agents and hardening the product. I'm guessing 30%-40%. Some might consider that a waste; but I don't. The software I'm creating nowadays is vastly more robust than I'd ever been able to create manually. I don't mean that the code is better. I mean the surrounding tests are vastly better. I have a higher degree of confidence than I ever had manually -- even when I used very disciplined TDD and Acceptance testing. And then there's the ability to quickly reorganize the modules and the architecture while keeping those robust tests running. That is a tremendous boon.

Louder for the ones in the back 🙂

@codewithantonio Happy to help🫶

@manupadev Finally, a technical limitation mentioned!

@codewithantonio With a cli you have to install it on the runtime the agent client lives in, with an mcp its just easy configuration to be done. Matters when you build for non technical users

cloudthing.gg 😏

Great slide from the “How to thrive as an AI-era developer” session at Google I/O today I think this T-shape will apply to not just developers but every job function We need to - go deeper with our domain expertise - go wider with adjacent skills and fields - learn to use AI well on top

Personal update: I've joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on it in time.

@mattpocockuk Wouldn't that many flags cause so many branches in the feature surfaces ?

Is anyone doing feature flag development with agents? Not tried it, but in theory feature flagging is an alternative model to PR's to getting work on main. 1. Put it on main, disabled by a flag 2. Deploy with the rest of the system 3. Unflag to selected users early 4. Fix bugs for those users 5. Unflag to more users 6. Repeat until shipped Feels like a perfect strategy to pair with agents

@EffectTS_ Once you see how your data flows through effect, you can't never go back

Same here 👀 Send us your best Effect quotes!

Vercel's AI-SDK is way better than Google's ADK

@ghostyfreak @conductor_build Hopefully @dotta has an answer ?

@manupadev @conductor_build 😭

I take it back. What they did to @conductor_build is unacceptable

It truly was the missing piece

This is misleading. This policy redefines the term "interactive" to mean "using an Anthropic front-end". If you use `claude -p` or Agent SDK to do something interactively, it now uses credits, not your subscription limits. So the "interactive use" heading saying "unchanged" subscriptions is not accurate.

@charlieholtz Let's keep improving Big Terminal mode and see how it goes. I have faith in you guys 🫶

since many people are getting tricked by Anthropic’s sugar coating, let me further clarify what the change really is programmatically invoking claude through “claude -p” (and equivalent) is how a lot of people have been using claude. this was an approved use case of the subscription quota since a year ago and people built various advanced tools and workflows with this the change today makes it so that such usage can no longer draw from the original subscription quota. instead, they now get charged at API pricing which can be over 25x more expensive than subscription quota if i keep using my claude subscription the same way, last month it cost me $200. next month it will cost me over $5500 - Anthropic will give me $200 credit back from the $5500 expense so I pay $5300 please don’t get tricked into thinking this is a “new free credit pool”. it’s taking back something that’s been part of the subscription value since the beginning