.@virtuals_io’s fair launch model for founders is actually pretty good. the incentives are well designed and aligned across founders, investors, and traders. that said, there are still quite a few weak points to work on. the biggest one is liquidity. LPs are way too thin. no serious person wants to buy size when it’s difficult to get in and out with size. and any large holder can effectively rug the chart if he sells irresponsibly. Improving liquidity depth should be a top priority. and i’m surprised it’s still not fixed. I personally like a few projects in the virtuals eco right now but the liq is holding me back from buying. also like every launchpad, mindshare is also critical. and the best way to generate mindshare is to produce massive runners. that’s exactly how the first virtuals season took off. one or two projects went on insane runs, and that kickstarted the entire virtuals meta. the rest was history.

🔥 World Cup season is about to go absolutely CRAZY 🌍⚽ The biggest stage in football is coming… dropping early alpha. @goldengoalsol is launching soon. This will be one of the biggest narratives this summer.

there is a billion dollar opportunity in creating a token standard for prediction market positions that plugs into DeFi liquidity pools and enables lending/borrowing against PM shares the core problem: PM positions go to zero at resolution so no lender touches them. a standard that wraps positions with risk metadata (probability, time to resolution, volatility) lets lending protocols price binary risk dynamically instead of avoiding it connect this to the wave of prop firms emerging to fund PM traders a lot of them will need infrastructure to extend capital against trader portfolios and manage binary resolution risk at scale that infrastructure doesn't exist yet and the demand side is already forming. the supply side is wide open. putting this out there for whoever wants to build it

$U1 Migration Update — Solana → Base 2,368,368 / 4,000,000 $U1 migrated. 1,631,631 remaining. The migration is moving smoothly and we're well past the halfway mark. Source verification is live on-chain via Drift, and the 72-hour security delay before claim continues to protect holders throughout the process. Window closes November 2026, or once 4M tokens migrate — whichever comes first. Next PR will bring full mobile support to the migration flow, making the process accessible from any device. If you've been waiting for that to migrate from mobile, it's coming soon. If you haven't migrated yet, the path is clear: 1:1 conversion. Solana → Base. Same project, better infrastructure, deeper liquidity, and the future of $U1. Migrate at the official Umbrae site. Support questions go through our Discord. discord.gg/nwayFvjmzm

DLMM infrastructure on Base is worth watching. liquidity mechanics matter more than most realize

iso_core is now fully built, compiled, deployed on a local test validator AND finally... all tests are GREEN! What did this test just prove? That the program completely works now! This is MONUMENTAL— from conception to execution, the Isometric model is now provably working! This core test proved: 1. The program can successfully create a market account on-chain with pertinent settings (bucket counts, price ranges, liquidity parameters, collateralization, oracle feed usage, expirations, USDC vaults, descriptions, market IDs, settlements and market quantities. 2. Proper invalidation for zero liquidity parameters and range inconsistencies. 3. Oracle update instructions work with only the authority being able to call it. 4. Only the authority can update liquidity parameters— liquidity instructions are fully functional. 5. The big one: math accuracy! Gaussian payout matches references— the implementation of Gaussian kernels gives sensible results for ranges and center-of-ranges 6. LMSR cost increases monotonically! The implementation of the LMSR formulas work as intended in all scenarios (buying more = costs more, scaled; LMSR marginal costs are positive and bounded— individual trades have a reasonable cost relative to liquidity parameters). Other relevant tests are now being conducted for positioning, additional settlements, multi-position ranges, full-range positioning and a few edge-cases such as minimum-width ranges and boundary-based positions. We are getting incredibly close to rolling out an entirely new primitive to devnet and cannot wait!

Our arbitrage bot is live on Base mainnet. Not a demo. Not a testnet. Real capital. Real trades. Real profit. All verifiable on-chain. This was a proof of concept — limited capital, limited pairs — to prove the system works before we scale. It works. What it does: - Scans 5 DEXes simultaneously — Uniswap V3, PancakeSwap V3, Aerodrome, SushiSwap, and Umbrae DLMM - Evaluates 20 cross-venue arbitrage routes every 3 seconds - Buys where the price is low, sells where it's high — fully automated - Auto-rebalances its own portfolio to stay capital-efficient Results from the proof of concept: - 700 on-chain transactions executed - Consistently profitable across every session - Every trade generates volume and fees for Umbrae LPs - Prices across our pools stay tight with the broader market This isn't just a bot. It's protocol infrastructure. Volume. Fees. Price accuracy. All automated. Now here's what matters if you're holding: Ignis Elite NFT holders will receive 30% of the arb bot's profits, distributed proportionally. The bot makes money — you make money. No staking. No lockups. Just hold and earn. What's next: - Scaling serious capital into the bot - cbBTC and new trading pairs coming - Reaching out to market makers to bring deep liquidity to the platform - We have a plan to accelerate the $U1 token migration — details coming once we finalize. This changes the timeline significantly. To any market makers reading this — our DLMM contracts are EVM-native. Live on Base today, but they deploy to any EVM chain. Arbitrum. Optimism. Ethereum. BNB Chain. Avalanche. We're building the first multichain DLMM platform and we're looking for partners who move real volume. DMs are open. Track the bot wallet yourself. Every trade on-chain. Fully transparent: basescan.org/address/0x1032D… The proof of concept is done. Now we scale. @Umbrae_Ignis

Our arbitrage bot is live on Base mainnet. Not a demo. Not a testnet. Real capital. Real trades. Real profit. All verifiable on-chain. This was a proof of concept — limited capital, limited pairs — to prove the system works before we scale. It works. What it does: - Scans 5 DEXes simultaneously — Uniswap V3, PancakeSwap V3, Aerodrome, SushiSwap, and Umbrae DLMM - Evaluates 20 cross-venue arbitrage routes every 3 seconds - Buys where the price is low, sells where it's high — fully automated - Auto-rebalances its own portfolio to stay capital-efficient Results from the proof of concept: - 700 on-chain transactions executed - Consistently profitable across every session - Every trade generates volume and fees for Umbrae LPs - Prices across our pools stay tight with the broader market This isn't just a bot. It's protocol infrastructure. Volume. Fees. Price accuracy. All automated. Now here's what matters if you're holding: Ignis Elite NFT holders will receive 30% of the arb bot's profits, distributed proportionally. The bot makes money — you make money. No staking. No lockups. Just hold and earn. What's next: - Scaling serious capital into the bot - cbBTC and new trading pairs coming - Reaching out to market makers to bring deep liquidity to the platform - We have a plan to accelerate the $U1 token migration — details coming once we finalize. This changes the timeline significantly. To any market makers reading this — our DLMM contracts are EVM-native. Live on Base today, but they deploy to any EVM chain. Arbitrum. Optimism. Ethereum. BNB Chain. Avalanche. We're building the first multichain DLMM platform and we're looking for partners who move real volume. DMs are open. Track the bot wallet yourself. Every trade on-chain. Fully transparent: basescan.org/address/0x1032D… The proof of concept is done. Now we scale. @Umbrae_Ignis

Today we're trying to build a new image for Local LLM using the Kimi model. Goal: reduce agent provisioning time so deployments become much faster.

The way to prove an AI agent is tied to a verified human owner with just a @VeryAI palm scan. Secure, private, and now available on Solana x.com/VeryAI/status/20187216…

This is the proof layer we needed @VeryAI @solana I just wired ClawKey‑verified OpenClaw agents into a Solana launchpad so agents can deploy trade tokens via API. Agent‑only token economy starts here.