My tweets are my own thoughts, so always DYOR.
Joined April 2023
- Tweets39.8K
- Following7.8K
- Followers75.2K
- Likes19.5K
Pinned Tweet
agents hiring agents sounds cool.
but who checks the work actually got done?
went through @termix_ai's live data to find out.
the simple version:
> you fund escrow
> the agent delivers
> you get one redo
> still unhappy? challenge it
> funds only move after that
termix never holds the money.
it sits in an escrow contract on bnb chain / base.
their dashboard says $29.7M escrowed and 573K jobs.
but that's their own dashboard, so I checked the listings.
holo card studio:
> 83 jobs
> 97 rep
> 99% pass rate
> $1 basic tier
$1 is the part I like.
you can test it yourself instead of trusting threads.
the risk is easy to spot too.
agent.family/scan shows 327K settled jobs.
the ones I scrolled were $20-75 each and tagged "optimistic".
meaning nobody challenged them.
some top earners show ~55 jobs, $30K+ settled, no stake listed.
could be fine. doesn't add up for me yet.
so:
rails look right.
volume still has to prove it's real.
🫴 try it: tell your agent "help me install termix.ai/skills"
(needs shell access + node 18)
nfa
if you are not on @creator_wire yet, you are seriously missing out
one of the best web3 agencies i have worked with so far
The once-a-day funding on @grvt_io's upcoming perps caught my attention.
Funding can get messy when you're trading different markets, so I’m curious to see how this works across stocks, FX and commodities.
Definitely one of the updates I’ll be checking out when it goes live. Not every change needs hype, sometimes the details matter more.
Most trading platforms focus on trading. @grvt_io seems to be looking at what happens to your capital when you're not trading.
With Grvt Earn and yield-bearing margin planned for 2.0, the idea of earning while keeping funds useful for trading is worth watching.
Still waiting to see how it all works in practice, but this is the part of the roadmap that caught my attention.
1.1M ethereum:0xad29f2723fcdbcf665f210f25e06f97477e417cf bought back and burned, nearly 1% of supply. Not just moved to a treasury either, actually burned.
Nice to see something delivered instead of just promised.
More interested in the Yield as Margin feature coming with @grvt_io 2.0 though. That could be genuinely useful.
One thing I find interesting about @primus_labs is its focus on builders.
Its Dev Hub gives developers tools to create data templates and integrate zkTLS into their apps. Less theory, more things people can actually build.
Check out Primus XP: s.kaito.ai/zsWfCmg
all that yapping about @axisrobotics on Kaito was worth it after all
how much allocation did you get?
buyback is already done
1.1M ethereum:0xad29f2723fcdbcf665f210f25e06f97477e417cf bought on the open market and burned. no promises, just execution.
with monthly buybacks intended and GRVT 2.0 around the corner, this is an interesting first step.
The evolution of Real Estate on-chain is moving quickly.
First came fractional ownership.
Then tokenization.
Now comes lending and capital efficiency.
RNT Lend, through its collaboration with @Aave, enables supported RWA positions to be used as collateral while stablecoin suppliers can potentially earn additional yield from USDT or USDC liquidity.
$30.5M Market Size
$21.5M TVL
$9.5M Borrowed
This is the infrastructure needed for RWA to move beyond simple ownership.
🤝 Paid partnership
USDT suppliers in RNT Lend are earning 12% APY. That interest comes from borrowers.
They post tokenized real estate as collateral, borrow stablecoins and pay interest every block on Polygon. The difference between the borrow rate and the supply rate goes to the protocol reserve.
The rate is variable and depends on utilization. Right now the USDT market is around 95% borrowed.
Proof of Reserves shouldn’t be a trust exercise.
@primus_labs helps crypto projects verify their reserves with onchain proofs instead of asking everyone to take their word for it.
Less “trust me,” more proof.
s.kaito.ai/zsWfCmg
An AI agent creates a video. Another analyzes the data. A third handles distribution.
Now imagine those agents sourcing each other’s services and paying for the work automatically.
That’s where things get interesting with @termix_ai. agent.family could help turn isolated AI capabilities into a connected service economy.
Not just agents working for us, but agents creating opportunities for each other.
What if your onchain reputation was backed by real data, not just wallet activity?
@primus_labs uses zkTLS to verify information from the web, opening new possibilities for loyalty programs, reputation scoring, and Web3 identity.
Less guesswork. More verifiable credentials.
Your internet activity could be more useful onchain than you think.
s.kaito.ai/zsWfCmg
Robotics has hardware, models, and funding. But reliable real-world data is still a major missing piece.
@AxisRobotics is tackling that gap with a data engine built for Physical AI, while $AXIS supports the growing ecosystem.
The infrastructure behind smarter robots deserves more attention.
s.kaito.ai/csqN2kD
