Where does the 11% yield on @grvt_io High Yield Earn actually come from? The app shows the rate, so I looked into the sources: → 40% Payment receivables, paid by shoppers settling card bills. → 30% Invoice financing, paid by businesses settling invoices. → 20% GRVT’s own market making, earning from trading spreads. → 10% Cash reserves for redemptions, earning nothing. The current blended rate is around 10.9%, while 11% is a target, not a guarantee. Real cash flows, but real risks too. This is not risk free yield. DYOR.

When applying for a loan, you often have to share months of bank statements just to prove one thing: you can afford it. That’s a lot of private information for a simple yes or no. This is where @primus_labs and zkTLS get interesting. Instead of exposing all your data, you can prove a specific fact, like meeting an income threshold, without revealing the underlying information. Less data shared. More privacy. Verifiable proof. Sometimes, proving less is worth much more.

with most nft projects, everyone watches the floor. with @meetmotes, i’d be more interested in how many of the 20,000 agents actually get funded and used. the planned 1% fee depends on real agent activity, so the collection size alone doesn’t tell us much. what matters is whether people actually put these agents to work.

an agent directory can show you which agents are out there, but that doesn’t mean you know if they actually deliver. that’s what i find interesting about @termix_ai. the idea is to give agents a place to get real jobs, prove they’ve completed the work, build a reputation and get paid in usdc or usdt. having a list of agents is one thing. having a place where they can actually work and get paid is something else. that’s the part i’ll be watching.

one thing i like about @meetmotes is how the owner stays in control. the nft represents ownership, the usdc stays in the smart account, and the agent can only do what the owner has allowed. i think this is important for AI agents. i wouldn’t want to give an agent full control over my funds, but letting it handle certain things within limits makes a lot more sense. that’s the kind of setup i’d be comfortable exploring.

if you create content for crypto campaigns, you probably know the struggle. connecting accounts and sharing personal data just to prove that your activity is real. that’s where zkTLS gets interesting. it lets you prove things from websites without exposing your login credentials or handing over all your underlying data. @primus_labs has already explored this with kaito, verifying activity that happens outside the platform. creators should be able to prove what they’ve done without having to reveal everything about themselves. that’s a much better model.

picture this: an AI agent needs fresh data for a crypto fund. it hires another agent to collect it, pays in USDC, and releases the funds only after the data is delivered and verified. no human in the loop. that’s the idea behind the agent economy @termix_ai is building. the key is the infrastructure behind it: → onchain identity → escrow payments → delivery verification → onchain reputation https://t.co/HBkTP0VgWl is where this is meant to happen. still early, but this is one of the first setups where agents paying other agents actually makes sense to me.

took a deep dive into @primus_labs this weekend, and the problem they're working on makes a lot of sense let me explain big institutions want to use public blockchains, but everything on a public chain is visible, including positions, trades and strategies. no serious fund wants competitors watching what it does it also can't go fully dark, because it still has to prove things like reserves and customer checks zkTLS lets you prove that data from a website or an API is real without revealing the data. FHE lets you do calculations on encrypted data without decrypting it private where it needs to be, provable where it has to be. that's what institutions actually need

something i hadn't thought about with @grvt_io: they say they own the whole stack, from settlement to the matching engine to the risk engine that's what makes one balance that earns and trades at the same time realistic. on borrowed infrastructure you'd end up moving money between products it's a company claim and i'll judge it when the features are live. but it explains the roadmap better than most "2.0" announcements do

this part stood out when i read through grvt's plans: fixed income is a $160T market and most normal people have basically been locked out of it @grvt_io wants to bring a slice of it into a balance you hold yourself, backed by tokenized assets. earn planned for end of october, yield as margin in november target yield is up to 11%, not guaranteed, and the asset partners get announced at launch. that's the part i'm most curious about I will be watching who they actually name