This looks like this could be exactly what I'm investigating. Anyone affected, please contact me or open your ledger and send me pictures so we can confirm if there's a bug in there (pictures: ledger board on top, spy implant below).

Will its AI pull in enough trading fees to beat the drag from losses and the expense of supplying liquidity? AI can definitely take the grunt work off people’s hands. The real bar is whether it can actually lift returns for depositors. @Tide_lp operates liquidity vaults for tokenized stocks on @RobinhoodCrypto Chain. With the newest rollout, communities can spin up their own vaults, pick the parameters, and collect a slice of trading fees while Tide’s agent actively runs the positions. Let's dive in. The mechanics highlight both upside and risk: - Deposit → Users add USDG for vault shares. The agent swaps some USDG into the chosen stock token and adds both to a Uniswap LP. - Earn → The LP earns swap fees while price stays within the set range. - Manage → Tighter ranges boost efficiency but go out of range faster. Rebalancing restores coverage, but costs gas and swap fees. - Exit → Withdrawals reflect the current mix (USDG and/or stock tokens). Share value can still drop even with fee income. --- ➥ Are the incentives aligned? Creators can charge up to 30% of earned trading fees. With a 10% fee on $100 of trading income, the creator receives $8, Tide receives $2 and $90 stays in the vault. This gives creators a reason to attract deposits and sustain activity. However, their compensation depends on fees collected, even when the underlying holdings lose value ➥ What does the AI actually improve? Tide says its agent considers price, volatility and US market sessions when deciding whether to adjust or hold. - Its decision log separates intentions from execution and includes failed transactions. - Contract rules constrain cooldowns, range widths, swap slippage and price deviations. - These controls can limit certain mistakes; they don’t establish that the strategy beats a simpler liquidity position or holding the assets. ➥ Can it scale beyond an experiment? Automated liquidity management already exists through several protocols like @ArrakisFinance and @GammaStrategies but mainly run without AI. Tide’s proposed advantage combines community-created vaults, tokenized-stock markets and inspectable agent decisions. However, demand still depends on sustained trading volume and depositor confidence. On October 9, Tide’s dashboard reported eight vaults and roughly 425 USDG altogether. Which we expect will start growing from here. ➥ The strongest risks are: - Unaudited contracts and protocol administrative powers. - Thin liquidity, sharp stock-token price moves and poor agent decisions. - Estimated fee APRs that exclude gas and price losses, rather than forecasting total returns. --- ➥ My takeaway $TIDE provides a practical framework for handing off liquidity management while still being able to monitor what’s happening. That said, I’m staying guarded, for now it’s an early-stage product with modest TVL and unaudited contracts. This case becomes far more compelling once permissions are audited and results consistently beat well-defined benchmarks. Until then, the strongest proven edge is transparency; outperformance is still a hypothesis.

There have been a reports on X and Reddit of wallet-draining by Ledger users. I traced the theft addresses and identified inflows from more hundreds of victim wallets across several major blockchains, including Ethereum, TRON, and Bitcoin. Total losses $86M+ bc1qjqgwejnp8dc0x2938x9n9954hj97t82unx49dl TK6DWNpNe1w2iJRNFpU8aHdrPTATxvXT6C TBkcUMYC7CkTK99tkTnaStQVBastfrs9d9 TCGE3xp6YGRKXxDZiLfysgJW3f22KfMNsW 0x69c8f401cfc6cd40ac94691d6d7c48e3b7a47841 0x033636e45d519bebb7b5c2520ca6ce56fbdb4f7a 0x83aeac166f6832ae3500000a24510a95a052a599 bc1qqnkwurxs99xkx5t4yffqhq3u6qwy0qpjyujtm9 bc1qgqheemzla77pesl227hdtgf5ykz62d0zvld26n TSDWtuZ2pARUVz4v3PkL2hi3iXPjowAr5a

Will its AI pull in enough trading fees to beat the drag from losses and the expense of supplying liquidity? AI can definitely take the grunt work off people’s hands. The real bar is whether it can actually lift returns for depositors. @Tide_lp operates liquidity vaults for tokenized stocks on @RobinhoodCrypto Chain. With the newest rollout, communities can spin up their own vaults, pick the parameters, and collect a slice of trading fees while Tide’s agent actively runs the positions. Let's dive in. The mechanics highlight both upside and risk: - Deposit → Users add USDG for vault shares. The agent swaps some USDG into the chosen stock token and adds both to a Uniswap LP. - Earn → The LP earns swap fees while price stays within the set range. - Manage → Tighter ranges boost efficiency but go out of range faster. Rebalancing restores coverage, but costs gas and swap fees. - Exit → Withdrawals reflect the current mix (USDG and/or stock tokens). Share value can still drop even with fee income. --- ➥ Are the incentives aligned? Creators can charge up to 30% of earned trading fees. With a 10% fee on $100 of trading income, the creator receives $8, Tide receives $2 and $90 stays in the vault. This gives creators a reason to attract deposits and sustain activity. However, their compensation depends on fees collected, even when the underlying holdings lose value ➥ What does the AI actually improve? Tide says its agent considers price, volatility and US market sessions when deciding whether to adjust or hold. - Its decision log separates intentions from execution and includes failed transactions. - Contract rules constrain cooldowns, range widths, swap slippage and price deviations. - These controls can limit certain mistakes; they don’t establish that the strategy beats a simpler liquidity position or holding the assets. ➥ Can it scale beyond an experiment? Automated liquidity management already exists through several protocols like @ArrakisFinance and @GammaStrategies but mainly run without AI. Tide’s proposed advantage combines community-created vaults, tokenized-stock markets and inspectable agent decisions. However, demand still depends on sustained trading volume and depositor confidence. On October 9, Tide’s dashboard reported eight vaults and roughly 425 USDG altogether. Which we expect will start growing from here. ➥ The strongest risks are: - Unaudited contracts and protocol administrative powers. - Thin liquidity, sharp stock-token price moves and poor agent decisions. - Estimated fee APRs that exclude gas and price losses, rather than forecasting total returns. --- ➥ My takeaway $TIDE provides a practical framework for handing off liquidity management while still being able to monitor what’s happening. That said, I’m staying guarded, for now it’s an early-stage product with modest TVL and unaudited contracts. This case becomes far more compelling once permissions are audited and results consistently beat well-defined benchmarks. Until then, the strongest proven edge is transparency; outperformance is still a hypothesis.

Kudo now uses @PhalaNetwork for agent attestation. A Covenant NFT records what an agent promised. A Phala attestation proves what code is making that promise. Promise, with proof.

TL; DR on Agency. @TryAgency ties a coin’s AI budget directly to trading fees. That creates a built-in funding loop, but also a weakness: the agent can lose its income precisely when its community needs it most. - It starts with @Pumpfun on @Solana. - At launch, 100% of creator fees are permanently routed to AGENCY. - The launcher chooses the AI model and personality, then gives up control over treasury decisions. - The first $20 funds model credits. The next $80 splits equally between credits and the coin’s treasury allocation. - After $100, 15% of fees funds $AGENCY buybacks and burns; the remainder supports compute and investments. That makes token trading the primary economic engine. The more interesting engineering sits between the AI and the money. A model proposes an action. A deterministic policy engine checks balances and limits, reserves the budget, and passes the transaction to an isolated signer. The signer then independently simulates it before execution. Accounting updates only after on-chain confirmation. This separation is crucial because an AI’s reasoning can be manipulated. Restricting execution reduces what a compromised model can do. However, an allowed investment can still lose money. --- If we want to compare to other Agentic Launchpads: - @virtuals_io VIRTUAL-priced bonding-curve launchpad plus EconomyOS agent wallets and compute. ACP adds paid jobs via on-chain escrow, and agents can earn without issuing tokens. Key difference: revenue diversity. Virtuals lets customers pay agents for real work, while AgencyPad’s funding loop relies more on its token’s market activity. - @CreatorBid CreatorBid’s Trenches/Clawlett stack keeps humans in custody via Safe wallets, while Zodiac Roles limits what agents can do. AgencyPad removes launcher control and instead uses its own policy and signing services, less owner discretion, more standardization. The real advantage is reliable orchestration, accounting, and execution, not access to an AI model. --- AgencyPad feels like a compelling sandbox for autonomous treasuries bankrolled by trading flow. The case gets stronger if agents can generate lasting community utility without living off perpetual hype and churn. But like most launchpads, the story unravels if volume dries up, because the mechanism meant to rekindle demand is the first thing that stalls.

I think you should understand how big is this. @GSR_io is rolling out @hare_xyz with @turtledotxyz, embedding its credit underwriting and risk playbook directly into onchain vault infrastructure. The debut lineup includes: → Hare USD Earn for leading USD stablecoins → Hare Gold Earn for Paxos’s tokenized gold Both products built atop @aave, with @Paxos Labs helping power the gold vault. A $100M GSR backstop, largely structured as a multiyear credit line, seeds the vaults with day-one liquidity. My take is simple: As tokenization scales, the market will want more than tokenized assets, it’ll want yield and utility around them. Hare probably is the best product on that demand, delivered with institutional-grade management. Higher.

While many are focused on 2049, @trondao just moved its USDT position from Aave to @sparkfinance. As a result, USDT borrowing on Aave V3 (Ethereum) hit double digits on October 7, with utilisation around 96%. Most liquidity was already lent out, leaving little room for withdrawals. At last check, Ethereum USDT borrow APY was 4.86% on Aave vs. 4.05% on SparkLend, meaning Spark still offers cheaper credit. More details: Pretty interesting move.

Spot trading for Pons (PONS) will go live on 7 October 2026. The opening of our PONS-USD trading pair will begin later today if liquidity conditions are met, in regions where trading is supported.

IMD by @surfcoderepeat wants an idea to become a working financial application through a network of AI agents. Its bigger ambition is for those agents to earn enough to keep building, eventually supporting companies managed entirely by the network. That starts with two connected pieces: Identity MD NFTs organise the workforce, while the $IMD token connects payments and incentives. Each of the 2,000 NFTs represents one swarm seat* Holders connect a machine running the IMD daemon and supply AI inference, the computation behind answers and code. The NFT lets an agent contribute to the network, although owning one doesn’t guarantee profitable work. Customers bring that workforce an objective, such as building a vault and its website. The system checks the request and breaks it into implementation, testing, review and delivery with pricing is 0.5 IMD per request, paid on @Ethereum through x402. When that work produces a token-bearing launch, the agents can participate in its economics. - 2% of the new token’s supply goes to wallets with accepted launch work, - 8% is allocated per connected seat at admission. The requester controls the remaining 90% with their value depends on what gets built and whether anyone wants it. The same workforce can also answer oracle questions for applications. With chain evidence, agents propose calculations at fixed blocks, and the deployer reruns the agreed calculation before signing. A contract can verify that signature and the agreement count. POOL4 connects its trading activity to funding. When selling pushes IMD inventory above the protocol-owned pool’s cap, excess tokens and proportional ETH are removed. - 85% of that IMD is earmarked for burning - 4.5% for stakers - 6% for orchestrator bonding - 4.5% for NFT inference nodes Stakers receive sIMD shares with growing IMD backing. Bonding would convert reserved tokens into ETH for computation, but bonding and node payouts remain unfinished. As of October 5, swarm deployment remains Sepolia-only, while published work includes imdUSD documentation and Ember World audits. However, there are some concerns like shared AI mistakes, deployer-dependent signing and unaudited POOL4 contracts remain substantial risks, including documented owner powers to withdraw liquidity and sweep staked IMD. The encouraging part is that each useful application can give people another reason to use IMD and bring more work to the swarm. If that demand becomes recurring, it could help fund the agents that build the next application. There’s still execution ahead, but the foundation is taking shape: a workforce, a payment token and a growing set of experiments. Turning those experiments into products people keep using would bring IMD closer to its ambition of an economy that sustains its own builders. NFA. DYOR.