matthias 0️⃣

cornetto or brioche pnl is up

The proxy thing for the secrets is quite elegant although I really don’t see why the hell would you slack this ? Unless you’re heavily already dependant on slack as a company. IMO the right approach (which we use internally) is the opposite. Start from codex desktop app

We just shipped Hallmark. The underwriting framework that gates every ForgeYields allocation. Yield aggregators give you safe-list returns. Anonymous farms ask you to trust them. We took the third path: frontier yield with the rigor of structured credit. 🧵

propAMMs just dropped on @Starknet Here's an ELi5 of how they work TL;DR: better trading UX, better asset pricing. Same frontend. New liquidity engine.

@ceccon_me thxxxx

@0xevolve Congrats on the launch! Didn't connect it was you behind this.

secretly making markets more efficient as it should be

@0xCramJam @0DFinance @avnu_fi just swap on avnu

@0DFinance @avnu_fi when/where can we give it a try 👀

1/ We are launching 0dAMM with @avnu_fi. A new onchain market structure for STRK: tight, transparent, programmable liquidity built directly on Starknet.

@kaiynne Not entirely convinced, having ran through this, you often get disappointed by long agent runs. I do agree on the multiplayer part tho

Was the first fan of it, but Claude desktop app is bullshit full of bugs, and their new update just makes it unusable anyways so I’m basically done at this point with it. Codex ftw

@SandipanKundu42 @zeddotdev Yes

Hey @zeddotdev, a better Git integration is the last missing piece to the amazing piece of software you guys have built. The current experience is just poor. I would have built a git plugin myself, but there seems to be some restrictions on the plugin api side to build what I was planning for. Please, let's iron it out. 🙏

The execution speed is unmatched 👏

We've been building ZK infra at @HerodotusDev for 4 years. storage proofs, cross-chain verification, trust-minimized everything. the Aave exploit broke my brain a little, not because it was sophisticated because the signs were right there and nobody had a way to catch them systematically. So we vibecoded a thing. introducing Risklayer ↓

why do we have so many grifters now ?

@Nlacombe_ Need access

finally a good zk use case, excited for this

@DanDeFiEd Bro is becoming a coffeinfluencer

Italian efficiency when it comes to coffee should be studied. In Italy: - Walk into a bar and look at the guy - Un caffe - 30 seconds later it’s ready - Shoot it - Leave €1 - Walk out In the US: - Join a line - Wait - Order coffee - Answer 12 questions: Size? Milk? Roast? Sugar? Temperature? Colombia beans? Name? How do you spell it? - $12.34 - Ask for a 20% tip. Click 5 times on a ipad to have a custom tip - Tap phone - ask where to send the invoice - Wait again on a different line - Someone call a name that sounds similar to mine - get the coffee - too hot, can't drink it - finally at temperature taste like shit

@pacovilletard Looking forward to try whatever you guys build! Chads

We're coming out of stealth to announce our cyber defense research lab. We are exploring data and post-training techniques to build superhuman cyber defenders. Our mission is to make sure the West always wins. The last 3 months we've built an automated data pipeline to create training data from 80k CVEs (aka public vulnerabilities). Our next topic? Post training a model that's better at fixing all the vulnerabilities in your codebase. Like really fixing them. Not saying it's secure when there are still ways to exploit them. Here are the questions that keep us awake at night: How do you train a model to defend without improving its capabilities to attack? What's the right reward? How to measure the defense capabilities? How do you create synth training data that reproduces real systems? What kind of access do you give an ai cyber defender? How far can you trust it? If you know insanely good cyber experts (red team, blue team, CTF aficionados) or ML engineers (synth data generation and post-training models), send them my way. We need to make models far better at defending.

@z0age Wonder why it just doesn’t do that in the background, as a user it’s just the worst ux ever

so claude code with opus 4.7 has a new feature: it'll helpfully explain to you over & over that the file it's just read is not malware sometimes it'll even explain how it's ignoring prompt injection & it might be malware the future is now and it's shaped like norton antivirus

A few notes after integrating both read and write priority fees on HL: - Many have been wondering why priority fee revenue hasn't ramped faster. It's a combination of two things. Makers need time to integrate, but the bigger issue is the validators, they have to explicitly enable the gossip priority config to respect the read auction ordering. Not all validators have done this yet, which means winning the gossip auction today doesn't guarantee you actually get prioritized access to the mempool. Rn you might end up burning your precious HYPE for nothing - Many retail traders don't know this, but before this upgrade you needed to pay validators to get access to sentry peering or other kind of dedicated setups in order to have competitive latency as an API trader, which had a cost of a few tens of thousands of dollars per month depending on the provider. This upgrade directly targets these setups, and will internalize most of these fees into HYPE burning. It's hard to estimate the exact numbers, but I would say that this upgrade just added $500k-$1m monthly buying pressure on HYPE, with potential for write fees to grow much more. - This is a first step from the HL team to try to internalize more revenue from latency and infra investments that most HFT teams spend. This is something traditional exchanges have been trying to do for a while (IEX introduced a 350μs speed bump and refused to offer colocation entirely, NYSE/CME keep building bigger colo facilities to capture more of this spend), and the biggest risk is destroying your microstructure because the exchange becomes too expensive for makers, or too slow. But the risk is definetly worth the reward. BIS research estimates the HFT arms race extracts $5B/yr in global equities alone, or about 0.5bps on all volume. Growth mode assets on HL charge 0.45 - 0.9bps in taker fees, so if priority fees capture even a fraction of that 0.5bps equivalent, that's potentially doubling protocol revenue on these markets. I'd estimate monthly infra spending from HFT teams on HL today to be in the ballpark of $5-10M between sentry peering, server optimization, and brainpower. Priority fees are how the protocol captures a share of that. My bold take is that trading fees will compress over time as competition forces HL to stay in growth mode for HIP-3 markets. Priority fees won't since they're driven by the arms race itself, which grows with more volume. My guess is this revenue line becomes more than half of HL's revenue in a few years if they manage to capture more flow from TradFi.