Darrin

@minhash @alfondotnet Looks fantastic. How do docs get added to the registry?

stop using MCPs to give your agents documentation. mandex gives your AI agent the same docs — but local, version-pinned, and 25ms instead of 800ms. One download. Works offline. No API keys. No rate limits. No monthly bill. 100+ packages. 1,500+ versions. Free forever. 🔗👇

@Hi_Mrinal I’m just about to launch a site with a hacker news/reddit type social voting interface where the content is entirely from vetted engineering blogs! I’ll share with you when it launches and make sure to include these in it

I really like to read engineering blogs from companies whose product I admire/use the most in my day to day life, its the best way to understand how the engineering teams at the top tech companies function and how they build scalable systems. It also helps me build a strong intuition for system design interviews as I get to know each and every pros and cons of components I will be using. I would really advice everyone to invest at least 30 to 45 minutes of their day to read a new blog post everyday and note down what significantly new they learned today ### 1. Netflix (@netflix) 1.1. Microservices - Netflix Conductor : netflixtechblog.com/netflix-conduc… - The netflix api re-architecture : netflixtechblog.com/engineering-tr… - Load shedding : netflixtechblog.com/keeping-netfli… - Making api more resilient : netflixtechblog.com/keeping-netfli… 1.2. Streaming - Content Popularity for Open connect : netflixtechblog.com/content-popula… - Video encoding at scale : netflixtechblog.com/high-quality-v… - Data compression for large scale streaming : netflixtechblog.com/data-compressi… 1.3. Architecture - Creating scalable offers platform : netflixtechblog.com/growth-enginee… 1.4. Infra related - Extracting image metadata at scale : netflixtechblog.com/extracting-ima… ### 2. Uber (@Uber) 2.1. Architecture : - Service oriented : uber.com/en-IN/blog/ser… - Push Platform : uber.com/en-IN/blog/rea… - Multi tenancy architecture : uber.com/en-IN/blog/mul… 2.2. API: > Uber's api gateway : uber.com/en-IN/blog/arc… ### 3. Airbnb (@airbnb) - 3.1. Avoiding double payments in distributed payments system : medium.com/airbnb-enginee… - 3.2. Knowledge Access and retrieval at scale : medium.com/airbnb-enginee… - 3.3. Promotion and communication platform : medium.com/airbnb-enginee… ### 4. Hotstar (@HotstarReality) - 4.1. Re-Architecting apps for scale : blog.hotstar.com/re-architectin… - 4.2. Pubsub for 50M(scale) : blog.hotstar.com/building-pubsu… ### 5. Incident.io (@incident_io) - Debugging deadlocks in postgres : incident.io/blog/debugging… - Choosing right postgres indexes : incident.io/blog/choosing-… - Time, timezones and scheduling : incident.io/hubs/building-… - data habits help build a data culture : incident.io/blog/how-data-… ### 6. Tailscale (@Tailscale ) - Perfomance : tailscale.com/blog/peer-rela… - Tailscale GitHub Action v4 : tailscale.com/blog/github-ac… - identity provider for scale : tailscale.com/blog/building-… - Better auth : tailscale.com/blog/workload-… - NAT learnings : tailscale.com/blog/nat-trave… ### 7. Discord () - How discord resizes 150 million images with go and c++ : discord.com/blog/how-disco… - How discord indexes trillions of messages : discord.com/blog/how-disco… ### 8. Twitter - Kafka as storage system : blog.x.com/engineering/en… - Twitter metricsdb : blog.x.com/engineering/en… - Storing and retrieving millions of ad impressions per second at scale : blog.x.com/engineering/en… - Distributed log : blog.x.com/engineering/en… ### 9. Dropbox (@Dropbox) - Infrastructure : dropbox.tech/infrastructure… - Intelligent DNS based load balancing at Dropbox : dropbox.tech/infrastructure… - Cross shard transactions at 10 million requests per second : dropbox.tech/infrastructure… - Improving the performance of full text search : dropbox.tech/infrastructure…

@gmickel @RepoPrompt Sign me up

The original Flow-Next was Claude Code workers + GPT 5.2 High reviewers via @RepoPrompt. The evolution doesn't care what you run: → Claude workers, Codex reviewers → Gemini orchestrator, local model workers, cursor-cli for implementation → Frontier models for complex tasks, cheap models for boilerplate Anything is possible. Full dependency graph means the orchestrator only dispatches what's actually unblocked. You see the graph; you control which model handles which task. Your swarm, your stack. Like, Follow and Reply if you want early access!

@ryancarson @salesforce @outreach_io What is that a screenshot of? A Ralph loop app of some sort?

Built my own version of @salesforce and @outreach_io. Just doesn't make sense to try to integrate with most external SaaS platforms now. I own all the data in pg and I can build exactly what I need. Yes, I can hear the comments already: "tech debt!" "slop!" "buying is better than building!" But I can iterate so fast now with Amp, it makes dealing with external apps seem like the ice age. Also, I can setup all sorts of self-improvement loops and scripts that I just couldn't do before.

@Dimillian What’d ya use to land on that design/theme? It looks fantastic

Codex Monitor is fully open source. I started it this morning, but I'm already using it for building the app itself! It's a macOS (Tauri) app built on top of the Codex App Server, which can manage all your Codex agents in a minimalist and pretty UI. github.com/Dimillian/Code…

@v_computer This looks fantastic

Yep, the rumors are true. We are releasing the easiest way to build Web Apps and Mobile Apps with Claude Code from your browser (with custom skills) next week. Comment below for credits + early access. Then Check DM.

@serafimcloud Looks killer would love to try it

As a visual person, I hate working in the CLI. But after using Claude Code a few times, it’s impossible to go back. The problem isn’t Claude. It’s the interface around it. So we built the best Claude Code client we could imagine. Runs fully in the browser. Parallel branches and projects. Live previews for every change. A calm, clean UI. Free to use with your own Claude keys. Reply if you want early access 👀

@voidmode @helloiamleonie Wouldn’t your clone instructions risk blowing out the user’s existing git repo? git clone github.com/andrefigueira/… your-project cd your-project rm -rf .git && git init

@helloiamleonie If you want to get starting with context engineering you could use my framework ;) github.com/andrefigueira/…

2023: Prompt engineering is a critical skill 2025: Context engineering is a critical strategy My colleagues just dropped a 41-page guide on context engineering. It covers: • Prompting techniques: Choosing the right words still matters • Query Augmentation & Retrieval: Finding the relevant pieces of information for the user query • Tools: Letting the agent interact with the outside world • Memory: Remembering past interactions Download it for free: weaviate.io/ebooks/the-con…

@skirano It would be killer to have a way to have claude-engineer simply tail a logfile for a dev server, vs full blown code execution, so that we could point it at any project's log file and have it react to errors that get logged. Like for a Rails app for example

Introducing Claude Engineer 2.0, with agents! 🚀 Biggest update yet with the addition of a code editor and code execution agents, and dynamic editing. When editing files (especially large ones), Engineer will direct a coding agent, and the agent will provide changes in batches. Batches are smartly selected based on file complexity. The code execution agent will run the code and check for issues. It can even start processes (like live servers) and end them. It's insanely powerful! 🔥

@ReadwiseReader Great will doc thank you

@iamdchuk Heyo! Haven't heard of this happening, if you could make an in-app bug report from the document you're having issues with, we'll be able to look at the logs for your account and see what's going on.

@ReadwiseReader I’m struggling with keeping my reading progress synced between an iPad and an android tablet, both running the app, both connected to wifi. I’m reading an epub, and whenever I make progress on one device, the other is stuck where I last left off. Help?

Well talk about a major whomp whomp to this process: Even with a pro account, I hit my Claude Sonnet limit in less than an hour. I guess it’s a good way to make sure I don’t spend 10 straight hours on any given thing :)

Might try using Cursor IDE today to see if that simplifies things

So far this is at least a very rapid iteration workflow. Ideally I’m not manually copying and pasting code around. Also, while the Artifacts are incredible, they don’t play nice with long source code files. Lots of “// existing code here” in the output which is annoying

Alright, goals for today: 1) Get the Product Requirements Generator module fully functional (give the app a feature idea, it iterates on the idea with you, then spits out a full requirements doc) 2) Get a simple landing page online to collect email signups for the future app

@adrianthedev Would you mind sharing some of the prompts you were trying to use? (Strip out anything sensitive of course). Also was this in a plain markdown repo or part of a rails or code project?

I had a tedious job: swapping every markdown notation for a custom component in all markdown files. There's still a lot to refine with these LLMs.

I got access to GitHub copilot and tried it out last night. I spent one hour to try and tell it what I want in a task I could have probably done in 15 minute and the task is still not done.

Tools that attempt to automate any coding project inherently lack nuance for the specific project type and have to rely only on the LLM to follow conventions. By narrowing to Rails, I can combine business logic guardrails with the LLM to boost code and project quality

More to come, I’m hoping to have it end to end functional by this weekend. If you’re a Rails dev interested in trying this out at some point, reply or message me and I’ll reach out. I truly think this tool can reduce handwritten code by 90%+, and avoid arcane code magic too :)

My goal is to get it stable enough to use it to recreate @dhh ‘s original blog in under 10 minutes video, but faster, without writing any actual code, and it generates clean rails code following conventions.

Also something worth calling out up front: yes I’m aware of all of the code AI projects, both command line feature builders and vscode extensions and cursor, etc. my theory here is an AI coding tool specialized in a specific opinionated framework will be way more effective

@KevinEspiritu Funny, I just posted this thread here after a long Twitter hiatus, also hammering along with Sonnet: x.com/iamdchuk/statu…

I'm working on my storytelling skills, so I trained a Claude 3.5 Sonnet model with a ton of different storytelling techniques and principles. I categorized them into general/longform/shortform techniques and started uploading transcripts of old videos into the model. I ask it to "check" the transcripts for violations of storytelling principles. It's shockingly effective at helping me see the gaps in my storytelling ability. I'll likely "pair write" with Claude for a while, as it does a lot of the blocking and tackling of basic structure, facts, etc. automatically. Crazy level-up as far as content creation goes.