Could be really cool for agentic defi. Nest it in an automation that reads broader signals, then automate frequent rebalancing with a particular strategy
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI?
Iβve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev
β’ 20-200x faster
β’ 40-400x cheaper (w/ output tokens free)
β’ Frontier composable intelligence optimized for decisions
AFAICT the shortest path to AI-based economic revolution
2
1
16
2,344
When you want this but self custody use wallet.ekubo.org
LP stock stokens on Base. Usually this is a complex process for people who have never done it before.
Now you can simply talk to the Bankr agent and it sets everything up for you. Agents are not only solving complex crypto UX, but now they are solving complex defi and making it more accessible.
In two years time everyone will be using an agent to manage their finances.
6
1
39
3,426
Agent wrote a tweet about another agent managing my stablecoin position
I don't pick ranges or when to rebalance, I just tell the bot to maximize yield after fees
4
33
1,896
Shouldn't be picking pools either. I just told it to pick which stablecoins to LP also from the claude app on my phone.
enabled by wallet.ekubo.org
22
355
We designed a way for anyone to enable the agent wallet to interact with any protocol. All you have to do is write (generate) an MCP server for your protocol(s). Teaching the agent to use a protocol is then as easy as adding your MCP server.
docs.ekubo.org/wallet/producβ¦
Excited to launch this because I've been using it to automate complex Ekubo position management for a couple weeks now. IMO agents obsolete a lot of vault products. They can be more patient and careful than people, and they don't have to give up MEV to searchers.
1
28
1,386
Moody π retweeted
π₯ going live with @sendmoodz (and co-host @clawdbotatg)
βοΈ he is an old friend and OG from early uniswap router design
π€ now he is a big time sloperator - come thru the slop computer! nitter.net/i/broadcasts/1qJVmyvqdβ¦
5
6
32
2,165
Excited to launch this because I've been using it to automate complex Ekubo position management for a couple weeks now. IMO agents obsolete a lot of vault products. They can be more patient and careful than people, and they don't have to give up MEV to searchers.
Introducing Ekubo Wallet: a native desktop wallet for Ethereum and EVM chains, built for doing onchain work alongside your favorite agent.
wallet.ekubo.org
5
4
52
6,736
The agent-first wallet we're building is starting to come together. I've always wanted to build a wallet. I will probably never use another wallet.
3
4
47
2,021
Cool, and a step in the right direction. Similar to what we're building.
But look at the choices it gives you. Hold your own keys and get zero policy enforcement, and the the keys are stored on disk optionally encrypted. Or get the guardrails and let them hold the keys, with no export command in the CLI.
There's no integration path for other protocols either. The applications they've wrapped get first-class commands and may come with a service fee. For everything else your agent has to assemble the calldata on its own, which is typically the job of the application.
We're building the version that doesn't make you choose. No service in the middle, a flexible but understandable policy engine, a standard for wallet-protocol communication, and we do a better job of keeping your keys secure.
Coming soon and it will be open source.
3
3
30
2,032
When I explain Ekubo Extensions, I'm often asked how they're different from Uniswap v4 hooks. Extensions are simpler and more powerful. Building a hook may get you attention, but Ekubo extensions enable you to build useful products.
docs.ekubo.org/concepts/exteβ¦
4
5
56
16,842
Tell your agent "give me exposure to nvda and aapl, and provide liquidity to the best pools on ekubo"
"rebalance all my positions around the current price based on volatility"
"withdraw my veSTONX yield, swap fees to STONX, and reinvest"
coming soon
2
38
2,036
This 63-call multicall rebalanced our staked STONX in a single transaction, and was constructed by Codex assisted by the Ekubo MCP server
robinhoodchain.blockscout.coβ¦
Our MCP server suggests a target allocation based on Robinhood Chain data queried in Dune--also written by Codex. It would be trivial to automate allocation based on different strategies
The best DeFi UI is no UI
3
1
31
7,193
STONX is a great experiment in AMM tokenomics
Both models w/ the same integrations, same codebase, even on top of the same contract (minimized fragmentation costs); which tokenomics do users prefer and which one provides better execution?
UNI vs AERO, but on top of Ekubo
6
2
32
2,499
Currently the best yield opportunities on Robinhood Chain: provide stock token liquidity on the new MetaDEX from @EkuboProtocol
9
4
61
4,395
Moody π retweeted
A new MetaDEX extension for Ekubo Protocol built exclusively for Robinhood Chain is live!
Try it out here: ekubo.org/swap?chainId=4663
15
17
172
87,844
.@circle could easily fix this problem with the USDC contract by implementing expiring approvals as described in ERC-8255
It just happened again today, the loss here is already more than double
eips.ethereum.org/EIPS/eip-8β¦
BarnBridge (@Barn_Bridge) got drained for 770K USDC
The attacker never needed a bug in the code. They got the DAO to vote them the keys.
1β£ Background
BarnBridge SmartYield is a retired yield protocol. Its Compound USDC pool still had a live provider contract, a working DAO, and around 90 wallets that never revoked their old USDC approvals. All the ingredients were still on-chain, just unwatched.
2β£ The setup
1. The attacker buys BOND, the governance token, and stakes it in the DAO. In a DAO this dormant, that was enough voting power to pass a proposal on their own
2. They deploy an empty upgradeable proxy
3. They submit one governance proposal - hand pool control to that proxy.
Description: "migrate proxy implementation". During review it looks like control moving to a fresh, harmless contract
4. They vote for it, queue it, and wait out the timelock
3β£ The switch
Once the proposal executes, the DAO calls `setController` on the provider and points it at the attacker's proxy.
Only then does the attacker upgrade the proxy to the real drainer.
4β£ The drain
Now the attacker is the controller. Two functions finish the job.
`_takeUnderlying(from_, amount)` pulls USDC via `transferFrom` from any `from_`, gated only by "is the caller the controller".
They call it for every wallet that still had an approval and sweep all of it into the provider.
`transferFees()` has no access control at all. It reads the provider's full USDC balance and sends it to `controller.feesOwner()`. The attacker's controller returns their own wallet.
- 92 wallets drained
- 776,575 USDC out
5
2
27
3,314