# DeFi — X 热门讨论 (2026-09-20 10:05 UTC)
## @0xsunnhi (Sunnhi) · 09-20 08:57 · ♥82 ↻0 💬103 one thing about @agenticscredit is starting to make more sense to me
ACS is not supposed to live inside one app
builders can pull the score through an API
other DeFi apps can embed it directly
and the same trading history follows the agent across that system
that feels much bigger than just another leaderboard
the interesting part is making an agent’s track record portable https://x.com/0xsunnhi/status/2101596699319755258
## @bgv1806 (Bùi Vương 🎯🐐) · 09-20 01:17 · ♥71 ↻0 💬72 good evening CT
the more i look at @Starknet , the less interesting the “BTC bridge” story becomes.
the bigger question is what happens after Bitcoin arrives.
strkBTC gives BTC a programmable form on Starknet, but the interesting part is the stack built around it.
you can move BTC into Starknet, use it across DeFi, provide liquidity, lend, borrow or trade.
then there is the part that most BTCFi products usually treat separately: privacy.
STRK20 lets strkBTC move between public and shielded states, so privacy becomes an option inside the same asset rather than forcing users into a completely different system.
and this is becoming more than a concept.
Xverse already supports the strkBTC flow, including bridging and shielding, while Endur supports shielded BTC staking and liquid staking.
that creates a very different capital path:
BTC → strkBTC → shield when needed → stake or deploy into DeFi → keep the position productive
the important word here is optional.
BTC is not private by default on Starknet.
you choose when to shield, when to unshield, and where to use the asset.
that matters because privacy without liquidity is difficult to use, while liquidity without privacy exposes everything.
Starknet is trying to bring both into the same execution environment.
and STRK20 is becoming broader than Bitcoin too.
it is now a framework for bringing privacy to ERC-20 assets, with the SDK and wallet APIs opening the door for more applications to integrate the same infrastructure.
so i don't see strkBTC as simply another wrapped BTC.
the more interesting thesis is that Bitcoin liquidity becomes programmable capital, while privacy becomes another property that applications can build around.
the next question is no longer whether BTC can move onto Starknet.
it's whether developers can turn that BTC into an entire financial layer.
that's the part i'm watching.
https://t.co/NmPfsABGFu https://x.com/bgv1806/status/2101480949976649869
## @FlareDevHub (Flare Devs) · 09-20 01:22 · ♥71 ↻16 💬0 XLS-66 and XRP-collateralized DeFi: two different credit primitives for developers.
XLS-66 is designed around borrower credit.
Liquidity sits in XRPL Single Asset Vaults, loans are fixed-term, underwriting happens offchain, and first-loss capital helps absorb defaults. Building on it means thinking about borrowers, underwriting standards, repayment schedules, loan brokers and default risk.
Collateralized DeFi takes a different approach.
XRP becomes FXRP, FXRP is deposited into a lending market, and borrowing capacity is determined by the value of the collateral. That requires reliable pricing, liquidity and liquidations rather than borrower underwriting.
These models complement each other.
XLS-66 can support underwritten credit where the identity and creditworthiness of the borrower matter.
FXRP + Smart Accounts can support permissionless XRP-backed credit where the collateral itself secures the debt, while XRPL remains the user control layer. https://x.com/FlareDevHub/status/2101481971780136963