cdnsteve

Checkout details on what this runs on for local AI here: https://tokenstead.ai/models/qwen3-6-27b

Local AI and open-weight models are becoming something to no longer ignore. I've started a community around this @tokenstead on X and tokenstead.ai YouTube and much more coming. DGX Spark on route, RTX 5090s and much more exciting builds. We need to have AI sovereignty!

Right, just announced https://x.com/BaiduAI_News/status/2069322806748410291

This feels like a bad precedent of things to come.

Is it just me or is the experience to get this up and running not feeling very Apple like? Is there no one liner to install or did I miss something?

1. Built an agent memory tool since all agents and clis are dumb and don't remember anything. Instead of prepping 300 project files and Md files I just say:

Check sugar memory for the latest thing we were working on.

2. The second thing is when making changes across a large codebase agents are also dumb at figuring this out and also grep 300 things, using tons of tokens. Instead I say

Check RemembrallMCP to analyze the impact of the change.

1. https://github.com/roboticforce/sugar

2. https://github.com/roboticforce/remembrallmcp

This is like a combination of a Sony Walkman meets Rasberry PI meets the prepper in me. Perfect.

And there it is, the vaccum of AI consuming your data everywhere, used to train their models all goes back to... ads.

Same things with OpenAI. Ads.

I feel like we're right back in the early 2000's Internet again at least they aren't popups, we hope.

But with these models being embedded into, literally everything, will your screen on your car start showing you ads before you can turn the AC on?

It's coming

Non profit files for IPO should be the headline.

So basically we are now waiting for a wave of massive new exploits coming via GitHub itself. Lovely

Small plug for what I built:

You need a code dependency graph: https://github.com/roboticforce/remembrallmcp Ask "what breaks if I change this?"

Saves 98% token usage. Saves 95% tools being called.

Runs as an MCP server, works for 8 languages.

It just works, you need to try it.

Currently, we face lots of churn in our ecosystems. API updates, CVEs that need updates, supply chain compromises - keeping up is getting harder and we need something to help us manage that, get notified and allow us to take action.

This is why I created ChangeSpec, an open standard to communicate software changes.

I'd love to get feedback and work on it together, it's still early but I hope you like it.

Does it help track me all the expenses from email and make them Booker ready or accountant ready. Worst paperwork job ever.

Working on a way for us to communicate software changes properly with people, systems and AI.

I'm frustrated at how we just do library updates and get whatever is next. Things break. Finding a changelog sucks. Vendor updates maybe put something in X.

We can do better so I'm starting an open initiative to tackle that.

By fall most execs will be asking about the new spend and GitHub alternatives will be researched.

Between 27x model costs and this, CVE exploits and downtime their platform is starting to feel like a questionable decision.

Run real War Room incident training drills with your teams, together, in a browser.

194 Engineers, DevOps, SRE's are already registered.

Start a free War Room. Invite your team. Compete in a countdown to solve the incident.

Learn about different kind of incident scenarios: - Kubernetes - AI - DNS - Database, networking and more being added weekly.

Multiplayer War Room plus new AI workstation scenarios are now live.

Blew through my usage in less than 1 hour after it was out. Max 20x plan. ouch

Are you using multiple agents on the same project? If you are, use git worktrees.

I personally prefer 3-4 concurrent terminal sessions, working on different projects. Claude code usually runs around everywhere with agents, and sometimes takes 20 minutes to get shit done. During that time I'm shipping open source work, etc.

I use my own tools to help: Sugar and RemembrallMCP. Sugar is used for memory that's stored outside sessions. So I say store this in sugar memory. If I open a new claude session later, and ask it to lookup sugar memory, it's there.

RemembrallMCP is a AST and code graph that helps the agents understand change impact of your codebase. Instead of insane greps veritical and horizontal, it uses this to quickly understand impact with very high accuracy, time reduction and low token limits.

Game changers.

Have them show you and talk about what they built. If they are not curious and building side projects, doing open source for excitement and making an app for a friend's business, their family member or some passion project, to me that's a red flag.

We are in a new age when the barrier to building is literally the lowest it's ever been. If your GitHub still is a dead graveyard fix that immediately.

I have over 2k commits in 2026 alone and I'm not even an IC anymore. Not including my work account.

I have like 12 side projects being built usually, working on 3-4 in parallel. A few are open source and growing.

All of this has helped me learn at a crazy rate. If your candidate is still just slogging 9-5, doing stand-ups then clocking out without anything extra.... That's all you need to know.

Curious, self learning, high potential people with a track record and good communication skills. That's it.

Is anyone using mail clients anymore?

Strap in, massive wave of security vulnerabilities incoming.

This looks to solve something I've been struggling with in my project, Sugar (1). Using the SDK and having sub agents running I found it difficult to have real- time insight into exactly what they were doing.

You can create a huge task list and Ralph mode can crank through it and also store persistent memory.

Interested in trying them together. 1. https://github.com/roboticforce/sugar

Cool project! This is also extremely useful to compare model bias across the board. There are some disturbing trends on certain topics.

Thanks for the feedback - we actually just shipped this. There's now an interactive security bug challenge right on the landing page. No signup, just click the lines you think have issues and check your answers.

You're right that the role is shifting toward AI orchestration. We'd argue that makes code review skills more important, not less - someone still has to catch the bugs in what the agents produce.

Hey nilo, how can I DM you? No contact details in profile or here.

YouBrokeProd

A fun simulation game of prod breaking where you have to race to troubleshoot and fix the issue. Great for SREs, devs devops, founders and teams to learn in a non prod environment but feels real.

We are in beta and would love feedback!

Was on show hm the other day https://news.ycombinator.com/item?id=47323915

https://www.youbrokeprod.com/

This is a great distinction. Right now all our scenarios are postmortem replay - you're dropped in after things have already gone wrong. But a pre-incident mode where you're reviewing a setup and have to spot the misconfiguration before it blows up is a really interesting idea. Different skill, equally important. Adding it to the roadmap.

Thanks! SadServers is great - love what Fernando built there. The main twist here is that the scenarios are built from real postmortems rather than generic server puzzles. The terraform one is modeled directly on the Claude Code incident from last week.

Lots more to come

thanks, let me know if you try a scenario

Thanks for trying it out! Just pushed a fix - there was a bug where the game engine wasn't starting properly after clicking GO. Should work now. Create a free account and give it another shot, would love to hear how you do.