Akash Kakkar

[Brad Traversy @traversymedia]

Learn software architecture. I know actually learning to code is hard and takes a long time. Learn architecture at a higher level. Also, don't just send multiple agents to work on multiple features. Stay with the agent and build feature by feature so that you can always follow the spec and know what is happening.

[Pi Changelog @PiChangelog]

Pi v1.4.0 is out!

Highlights:
- codemode models.classify() now accepts images for Jev and Clef models on typesafe, cloudflare-workers-ai, and OpenRouter
- Compaction and branch summaries now use the model's output limit, fixing failures when reasoning used most of the fixed summary output limit
- Codemode scripts queue for a free worker instead of failing; PI_CODEMODE_MAX_WORKERS raises the limit
- Rex terminal support: OSC 8 hyperlinks, truecolor, and inline images

Complete details in thread ↓

[Akash Kakkar @akash_k07]

@unclebobmartin Uncle Bob, I'd really love to read about your AI workflow. have learnt a lot from your books, I'm sure your workflow for Agentic development will be very efficient and worth adapting.

[KRAFTON AI @Krafton_AI]

Today, we’re releasing a preview of Raon-1 mini, a 30B MoE LLM with 4B active parameters, trained from scratch by KRAFTON AI.

With our efficient architecture, it matches Muse Glimmer on coding/agentic tasks while being much faster and more memory-efficient! 🔥 Plus, natural, human-like conversation.

This is just a first look. The final release (and bigger models) are coming soon! 👀

👉 Try it now, free for a limited time: https://t.co/uSrLFIGPWy
🤗 Open weights + DFlash drafter (Apache 2.0): https://t.co/XWVmxsDRei

[Akash Kakkar @akash_k07]

Absolutely true. So tiring and exhausting.

[Armin Ronacher ⇌ @mitsuhiko]

@can1357 Worst part of everybody having access to AI is everybody being an expert on everything right now. Makes the Twitter discourse even more tiring.

[Matt Pocock @mattpocockuk]

Prompt of the day:

Create a pre-commit hook that prevents files over 5,000 tokens. Tokens = bytes/4.

Run only on staged files. Ignore files which are not authored by agents, like lockfiles or codegen.

Provide explicit instructions on how to split files in the error message.

The simplest investment I ever made in my codebase. So good.

[Armin Ronacher ⇌ @mitsuhiko]

🎙️ New Episode of State of Agentic Coding with @bentlegen and me is out. This time with @GeoffreyHuntley as guest.

- Code mode and self modifying harnesses
- @dhh's rails keynote
and a lot of other things

[Pi @pidotdev]

For your weekend reading: Build with Pi Durable on Cloudflare, a walkthrough of how to create a multiplayer chat agent, including subagents, on Durable Objects.

https://t.co/Y7EBwsV52x

[ggml @ggml_org]

llama.cpp v0.7.0

https://t.co/EwOsb6WKeC

[OpenCode @opencode]

Gemini 4 Argon is now available in OpenCode Go

$60 usage for $10 til Nov 30

[Mario Zechner @badlogicgames]

designing a new plugin system. everything leads back to pi-durable concepts.

i'm pi-durableing all the things now.

[Web Dev Simplified @DevSimplified]

I just uploaded a new video! (Vite+ Just Killed Every Major Config Tool) https://t.co/pHVYMpA5Hb

[Uncle Bob Martin @unclebobmartin]

Messy code slows agents down, just like it slows humans down. I've seen agents struggle over maintaining messy code, where every change they make breaks something else they hadn't anticipated. They can thrash for some time trying to deal with it.

So I think all the same principles apply; but with different thresholds. For example, in my own code I limit cyclomatic complexity to 4 or so. But I allow agents to go as high as 10. They seem to deal with it just fine. My hypothesis is that their short term memory is larger and more accurate than ours.

So I agree with you that the principles will need a different tuning for agents than for humans. But I am pretty sure that they still matter.

[Neelesh Misra @neeleshmisra]

The Supreme Court ordered that protesters not be detained. Delhi Police did not comply.
Delhi Police has been one of the foremost police forces of the country, one that we have admired, but it has shamed itself in the past few months with its vengeance-filled handling of public protests. This collapse of empathy is a failure of leadership and training.
Compare this with the way Mumbai Police and other state forces dealt with young protesters.
The illegal detention of peaceful youth (even passers-by who were not protesting) and especially journalists on duty, despite Supreme Court directives, is a blot on its image that the mushy tweets from its social handles cannot wipe.
India’s youth have very real concerns and anger over several issues that need to be addressed. They must be given space and platforms to express them. It’s their right. Violence by protesters must be punished always but crushing peaceful protesters like this can and will only boomerang.

[Token Harbor @TokenHarborAI]

Four free models. One API.

Try MiMo V2.6, DeepSeek V4.1 Flash, DeepSeek V4 Flash, and Qwen3.8 Flash on Token Harbor.

No credit card needed to get started.

[UN Human Rights @UNHumanRights]

#India: We are concerned by reports of mass detentions of demonstrators, protest leaders, members of civil society organisations, lawyers and journalists, before and during today’s large-scale demonstrations in Delhi.

We call on the authorities to ensure that the right to peaceful assembly is fully respected and fulfilled.

https://t.co/tqrQyZoXBJ

[OpenRouter @OpenRouter]

Step 5 Preview from @StepFun_ai is live on OpenRouter.

Their new flagship for agentic work: a sparse MoE (27B active, 600B total), 1M context, and text, image, and video input. Strong at coding and professional knowledge work, especially finance.

Try it: https://t.co/kQR1mciHSv

[StepFun @StepFun_ai]

Step 5 Preview now exists where you build.

Now live on @OpenRouter. A week of free access is rolling out today across @opencode, @cline , @NousResearch, @kilocode, and more.

Flagship-tier intelligence for agentic and professional work, at substantially lower task cost.

Keep your workflow. Switch your model.

Try Step 5 Preview on OpenRouter: https://t.co/Z1nOT0rDj8

[Elon Musk @elonmusk]

Thank you, Rahul.

This is indeed troubling.

[Unsloth AI @UnslothAI]

Windows now has sandboxing!
Microsoft released an open-source repo, mxc, for sandboxed code execution.

We collaborated with Windows to add mxc OS level sandboxing to Unsloth which adds just <100 ms of overhead.

GitHub: https://t.co/aZWYAtakBP
Guide: https://t.co/mhgPeh9Fld