backend / systems engineer | distributed systems | building @sendryxhq | payments & infrastructure | @superteamng | svm @solanaturbine | alumni @alx_africa
onchain
Joined December 2014
- Tweets14.5K
- Following691
- Followers1.7K
- Likes30.1K
Vector 🥷⚡️ retweeted
THIS IS ACTUAL F*CKING GOLD
Anthropic just shared one of the best tricks for Claude Code.
The idea: run Sonnet, Haiku and Opus 5.5 together, each one doing only the job it's best at.
Opus-level judgment on every big call, without paying Opus rates for every routine turn.
You just launch Claude Code like this:
claude --model sonnet --advisor opus
here's the split:
→ Sonnet 5.5 runs the session and writes the code
→ Haiku 5.5 explores files and docs in parallel
→ Opus 5.5 stays on call as advisor for the big calls
→ you stop spending Opus on work that doesn't need it
Opus only steps in at three moments:
→ before a plan locks
→ when the same error shows up twice
→ before calling it done
Sonnet builds. Haiku explores. Opus reviews.
JEV engineering is the same move one layer down: the forks that need no thinker (which file, which tool, retry or stop) go to JEV in milliseconds, and the big models only see the ones that split
the full tree
> Sonnet 5.5 runs the main session and edits code
> search maps the repo on Haiku 5.5
> grep finds call sites on Haiku 5.5
> docs pulls the API reference on Haiku 5.5
> Opus 5.5 on call as advisor, reads the whole transcript when consulted
> JEV routes micro-forks in milliseconds
paste the tree and this prompt into Claude Code ↓
"Rebuild my Claude Code setup around this tree:
1. Run the main session on Sonnet 5.5 with Opus 5.5 as advisor.
2. Create three read-only Haiku 5.5 subagents: search (maps the repo), grep (finds call sites), docs (pulls the API reference).
3. Add one rule to CLAUDE.md: consult the advisor before locking a plan, after the same error twice, and before calling a task done.
4. Route fast deterministic forks (tool choice, path selection, retries) through JEV.
Show me every change as a diff first. No edits until I say go."
↳ code.claude.com/docs/en/advi…
save it, you'll use this every single day.
Vector 🥷⚡️ retweeted
. @solana has opened so many doors for people, myself included.
Now I'm in a position to help open a few more.
Happy to join @SuperteamTalent as a Talent Scout.
If you've got the skills and you're looking for an opportunity in the ecosystem, reach out.
Let's make something happen.
Vector 🥷⚡️ retweeted
So basically, today PHP got two critical bugs that let anyone run code on your server without even logging in, which is bad, and then the details leaked before the patch was even out, which is worse, and it works on WordPress and MediaWiki and Magento and Joomla and Laravel, and Censys says like 1.5 million servers are running a version with no fix, so the NSA was like "take your system offline," and the FBI was like "take your system offline," and CISA and the UK and CERT-EU were all like "take your system offline," and then the Prime Minister of Japan, the actual Prime Minister, posted at 5:55 in the morning telling 48.8 million people to take their system offline, so everyone took their system offline, like the Dutch cybersecurity agency took its own website offline, and the Dutch parliament went offline, and the Port of Rotterdam went offline, and Finland and France and Denmark started pulling government sites, and Wikipedia went down, and Archive.org went down, and Asana and GOG and Something Awful went down, and the University of Vienna's webmail went down for some reason, and that's just PHP. Because GitLab also dropped a CVSS 10.0, which is the highest score they give out, for a bug that's been there since 2014, so that's every self-hosted GitLab for twelve years, and then LiteLLM 1.107.0 got hijacked and shipped a credential stealer and a remote access trojan in every single wheel and the Docker images, including latest, and you didn't even have to import it, just installing it was enough, and it's the second time this year, and then Railway paused every deployment on the entire platform because someone got into the environment variables, so now everyone has to rotate every secret they've ever stored there, and TiDB Cloud said someone was walking around inside with stolen credentials for basically a month and an extortion group says they took 10.6 terabytes, and Zeabur got breached for the second time in two months and is just shutting down, like the whole company, and Node.js patched remote code execution and a TLS bypass, and CISA says Plex is being actively exploited, and someone took over 300 Tor exit relays to steal SSH passwords, and SafePay ransomware has hit over 600 critical infrastructure orgs, and India's biggest bank filed a data loss incident, and Japan has been in an official state of emergency in cyberspace since October 9, which honestly explains the 5:55 AM post. Oh, and eMudhra, a certificate authority that's issued over 60 million certificates, whose entire job is proving websites are who they say they are, was supposed to be using private keys that are random numbers over a hundred digits long, and one of them was 4. Just 4. And the others were like 66 and 237. Anyway, it's Wednesday.
Vector 🥷⚡️ retweeted
Introducing tsrs: a Rust port of the TypeScript compiler 🦀
We are not forking TypeScript!
◆ In close collaborating with the TypeScript team
◆ Up to 8x faster than TS 7.0 & ~70% less memory
◆ Runs in WASM, so it works in your browser & on Workers
◆ Soon part of the Oxc toolchain
npx tsrs
blog.cloudflare.com/typescri…
Vector 🥷⚡️ retweeted
Excited to share that Evolution has joined Nuvion @ournuvion 🚀
When we started Evolution, we had a simple vision: make moving money across borders easier and more accessible for individuals and businesses, enabling them to go Global from wherever they were.
We believed that accessing global financial services shouldn’t be complicated simply because of where you live or where your business operates.
What started as a platform for exchanging digital assets grew into something much bigger. We built infrastructure that enabled people and businesses to hold multiple currencies, receive international payments, move money across borders, and access financial tools that were previously out of reach.
Along the way, we served thousands of users, processed millions of dollars in transactions, and learned firsthand what it takes to build financial infrastructure in emerging markets.
But more importantly, we saw just how much work remains to be done.
As we continued building, one thing became increasingly clear: solving cross-border banking challenges at scale requires more than a great product. It requires strong financial infrastructure, regulatory reach, and the ability to serve businesses wherever they operate.
That’s what makes this next chapter with Nuvion particularly exciting.
Nuvion is building an AI Powered Global banking platform for ambitious businesses, bringing multicurrency accounts, international payments, FX, virtual cards, and stablecoin infrastructure together in one place.
Our visions have always been closely aligned, and joining forces gives us the opportunity to pursue what we started at Evolution on a much larger scale.
For me, this is an opportunity to take everything we’ve learned building Evolution and apply it to something with the infrastructure, reach, and ambition to serve businesses around the world.
I’m especially excited that members of the Evolution team are continuing this journey with me at Nuvion. We’ve spent years building, solving difficult problems, and learning together. Now we get to bring that experience into a much bigger mission.
To everyone who believed in Evolution, our early customers, investors, partners, community, and the incredible people who contributed to our journey: thank you.
You trusted us when we were still figuring things out, challenged us to build better, and helped turn an idea into a real business.
I’m incredibly proud of what we built, and even more excited about what comes next.
Evolution was never just about building another fintech company. It was about making global banking accessible for the people and businesses who need it.
That mission continues at Nuvion.
The Job is just starting
Vector 🥷⚡️ retweeted
LayerFS is a Rust-powered filesystem for AI agents, enabling zero-copy workspace branching, deduplicated snapshots, parallel rollouts, and efficient state recovery.
ALT Image: LayerFS is a Rust-powered filesystem for AI agents, enabling zero-copy workspace branching, deduplicated snapshots, parallel rollouts, and efficient state recovery.
Vector 🥷⚡️ retweeted
Solana just halved its target slot time.
400ms → 200ms
2.5 → 5 target slots per second
1.6s → 0.8s validator leader windows
For perspective, Ethereum still runs 12-second slots.
That's 60x longer per slot than Solana's new target.
Solana's first full epoch averaged 218.3ms.
We're now looking at more frequent transaction executions, quicker feedback, fresher onchain data for defi..
That's just block production btw.
Alpenglow targets ~150ms finality. Not live yet.
Solana will be the home of onchain finance.
Vector 🥷⚡️ retweeted
Y'all aren't thinking about ai productivity correctly. I have one customer support guy at boot dev, he can handle the entire support load (low 8 figure revenue) in just a few hours a day
Theres literally no business value to me in automating even 80% of his work, because I'll still need him for that last 20%
HOWEVER he's not online 24/7. There are a subset of tickets, that, with a really good model, we could give answers to that are no worse than his, nearly instantly.
inj other words, you need to be more focused on using AI to improve products and services - and less concerned with cutting marginal costs.
Vector 🥷⚡️ retweeted
The @excalidraw mcp is sick.
Have it live draw explanations (with pencil sounds and all) then pop open to the app and make your own changes.
Learning with ai is so fun. (Also, don't stop learning, your brain is too valuable)
Vector 🥷⚡️ retweeted
ive been leveraging ai to teach me topics in ways i like to learn
built a memory simulator: pick the hardware, OS (Linux, macOS, Windows) and apps, then watch what happens when you run out of RAM
here: an 11 GB game on a 16 GB machine pushing Photoshop and Slack's idle memory out to swap
Vector 🥷⚡️ retweeted
Rust has one serialization library that is Serde. Over the years I collected a bucket list of all the things that I do not enjoy about it. Finally did something about it and finished Deser. Here is a small summary of the why, what and how: lucumr.pocoo.org/2026/9/29/d…
Vector 🥷⚡️ retweeted
I made a video about this but had to delete it. Many of these people come with immense arrogance rather than openness to learn.
It's sickening. Everybody wants to prove they know it all by putting the other person down. Chill bro, you don't work at Jane Street, on flight control systems, or at Google or Cloudflare.
It's frustrating to see people without an open mind to learning and improving.
Vector 🥷⚡️ retweeted
I think continuous profiling should become a default part of observability in the AI coding era.
AI can generate a lot of correct code very quickly, but correct code is not always efficient code.
An agent might choose an unnecessarily heavy library, allocate too much memory, repeat expensive work, use a poor algorithm, or simply implement something in a way that costs far more CPU than necessary.
Metrics can tell you that CPU usage or latency is getting worse. A flame graph can show you where the program is actually spending its time.
From there, you can inspect the hot paths, review the implementation manually, research better approaches, replace inefficient libraries, or use an AI agent to investigate possible optimizations.
As AI makes writing code cheaper measuring the actual performance of that code becomes even more important.
