I built reflect: a journal that asks you one good question a day.
The catch: the LLM runs entirely on your machine. No account, no network, no upload. Your diary literally never leaves the device (there's a test that blocks every socket to prove it).
Powered by @nobodywho_ai (llama.cpp).
⭐ Link in the comments 👇
Repo (free & open source): github.com/nikolovv861/refle…
Runs on Qwen3-4B locally. If you build local-AI apps, the engine it's built on is worth a star too: github.com/nobodywho-ooo/nob…
EchoKly retweeted
Audit complete for @TheIndexFi's Baskets - tokenized baskets built on Uniswap v4, coming to the @RobinhoodCrypto chain.
1 high, 11 medium, all resolved before launch.
Kudos to the team for taking security seriously.
SBSecurity is now Index’s official security partner. 🤝
EchoKly retweeted
Dear builders and auditors, you don’t have to spend all your tokens on Claude anymore.
Introducing Kann AI.
The first open-source AI model fine-tuned specifically for finding vulnerabilities.
Free to use, built for Solidity smart contract security.
Lfg
@ParlayItGG is now LIVE!
Prediction markets just got a massive upgrade - native parlays, on-chain settlement.
Stack your predictions. Multiply your payouts.
"You're not a dev, why are you at a hackathon?"
Just took 1st place at EPFL.
55 projects. 36h. $16K+ in prizes.
My 7-step playbook for non-devs 👇
Huge Thanks @xrpl_commons & @bsaepfl 🙏
nitter.cf/echo_kly/status/203718…
EchoKly retweeted
Stani lays out the abundance thesis clearly.
The numbers already proved it. The world added 602 GW of solar in 2024, then blew past that pace in 2025 with 380 GW installed in the first half alone. China hit 1.2 TW of installed solar by year's end.
Where it gets interesting for us is the gap he identifies in how this gets financed. Most solar growth is utility-scale, meaning it’s big enough to attract traditional infrastructure capital.
But commercial rooftops sitting on warehouses and factories across Europe are ideal for solar and are completely underserved.
The deal sizes are too small for conventional project finance to touch efficiently.
That's the exact problem tokenization solves when you apply it to assets that actually need new rails and not treasuries that already trade fine.
The RWA market has crossed $30B, but most of it hasn't gone near real infrastructure yet.
We're building for that layer.
IntelliShares tokenizes solar and battery storage deployed in Southern and Eastern European grids still running 40%+ fossil fuels, where every kWh displaces the most carbon. Investors earn from real energy output, not financial engineering.
Stani's right that the abundance era doesn't need more proof.
It needs capital infrastructure that moves at the same pace as deployment.
That's what IntelliShares is focused on.
If you're in BD or web3 marketing - stop scrolling
Hackathons have always been "for devs." Nobody told you that's changing.
First AI agent hackathon $500 prize
hackathon.leapility.com/
Stuff they don't want you to know.
Claude released its 4.6 Opus version, but in addition to that, there is some extra FREE usage you can claim (around €42)
Settings → Usage. It should be at the top.
EchoKly retweeted
🚨JUST IN: Pashov Audit Group🤝Discord
Welcome to our new community with technical topics around security, AI and dev. We live and breathe web3 technologies🫡
Join, RT and comment below for a chance to win 3x full-year Claude/GPT subscriptions👇
discord.gg/pashovauditgroup
EchoKly retweeted
Last year I wrote The Crypto BD Black Book.
It covers everything from ecosystem building to the frameworks I have used over the years growing crypto ecosystems.
Making the digital copy free for the community now.
Download, read, and share your thoughts!
docsend.com/view/y3s4be5fpv6…
5 years in the trenches. 15 months of grind with countless all-nighters. At least 10 moments I nearly quit.
I’ve read hundreds of great books, but this one feels different 😀
Thrilled to share my first book: The Crypto BD Black Book.
Written for operators and builders, it’s a playbook for surviving the grind and winning anyway.
EchoKly retweeted
i read + annotated the entire 300+ page Messari 2026 thesis on a long-haul flight so you didn’t have to…
my thoughts + takeaways:
> the entire thing took me about 3hrs to cover back to back (better than expected). My 3 fav essays were “crypto money”, “tradfi x crypto”, and “decentralized Internet finance”. Covers 80% of the trends that happened this year. Honorable mentions on the crypto x AI and DePIN pieces. there’s an entire encyclopedia of the top50 chains that you can skim, although there are some interesting ones like Tron/stablechains/Solana propAMMs thrown in.
> cryptomoney was a really great piece to open, discussing monetary premiums vs. actual value/fee accrual of the chain. TLDR BTC is accepted as a macro asset, ETH somewhat but actually premiums correlate with BTC rather than stand on own 2 feet. Recurring issue for L1s is the lack of take rates in a cheap-chain world. If we say L1s are like cities, chain revs are like city tax - every other chain these days has low tax rate, disincentives activity on something like ETH with high tax rate. But low tax rate just means no chain revenue, which forces monetary premiums, which can only accrue to very few assets.
> the report itself has contradictory views on ETH. It says rightly in “cryptomoney” that ETH is 4th in chain fees, making 17% overall, and fees overall dropping (bearish take). But in “chains encyclopedia” ETH entry it says ETH is winning “all the core sectors” eg. RWAs (justifies bullish take). So why low revs even with RWA dominance? IMO it might be RWAs have low velocity of money - more like a passive vault product rather than active trading product, so doesn’t build trading volume.
> the word that popped out throughout the report but especially in the TradFi x Crypto section was “distribution”, which imo has become a bit of a buzzword. Ofc distribution matters, but only if it’s tapped in to some existing product flow/flow of funds (eg. coinbase lending allows users to coveniently access Base Morpho vaults from Coinbase UI). This isn’t always clear in the report. Eg. Discussing PYUSD rise it cites “PayPal has distribution” but doesn’t say HOW PYUSD is integrated into flow of funds/product flow to tap into distribution.
> by the same token, TradFi report seems a bit overly bullish on consortium chains just because they have “distribution”. We’ve seen this before - LVMH/other luxury brands launched a consortium chain called Aura for NFTs last cycle, but how many ppl have heard of this? Real challenge is how they integrate consortium chains with (1) real world flows and (2) the outer world of DeFi
> remain partial to the DefiBank thesis. This cycle is not just about bank coming to crypto but also how crypto goes after banks with faster global permissionless rails. Unlocking save, earn, borrow for a global audience is perhaps the killer consumer app
