We’re building Geldux on @base. Not just another perp interface. The goal is a trading terminal that gives onchain traders more context before they make decisions: • Heatmaps • Liquidation data • Whale activity • Trader intelligence • Portfolio insights • Mobile-first trading UX The feedback has been clear: Base traders need products that feel fast, practical, and genuinely useful. That’s the direction for Geldux.

I’m Suhaib, building @Geldux. I started with one frustration: onchain trading gives you execution, but not enough info. So I’m building a @Base native perp terminal with Atlas risk intelligence, session trading, and cleaner market reads. Still early, but moving every day.

Hello @base 👋🏻

Atlas is the intelligence layer. Session trading is the execution layer. Together, they make Geldux feel closer to how an onchain trading terminal should work. A trader should be able to understand market conditions, check position health, see risk signals, and then act without fighting wallet friction every few seconds. The goal is not to hide risk. The goal is to make risk more visible while making the actual trading flow less painful. That is the part we are focused on improving. #DeFI #base #dex

Builder note #6 from Geldux: Over the last few days, most of the work has been focused on tightening the foundation instead of adding surface level noise. Geldux is now moving through the kind of stage where small product decisions matter a lot. The trading flow can look simple from the outside, but under it there are many things that need to behave correctly every time: collateral accounting, oracle checks, margin mode handling, position lifecycle, liquidation reads, portfolio state, history, and the way risk information is shown to the user. On the contract side, the clean Geldux Perps V2 deployment is live on Base Sepolia with gUSD test collateral and five curated markets. The current system supports both cross and isolated margin flows. We completed smoke tests for isolated and cross positions including opening, reading risk, partial close, full close, and cross collateral withdrawal. The most important part was the accounting result. Vault accounting stayed clean. No bad debt was created. Open interest returned to zero after the test flows. Cross collateral was fully withdrawn after the cross test. Risk reads were available through RiskLens. That gives us a stronger base before bringing more testers into the product. The product work has also been moving in parallel. A lot of focus went into separating global market intelligence from wallet specific data. This sounds like a small detail, but it is important. Market level information should not change just because one wallet connects. A connected wallet can add personal context, but it should not mutate the global view of the market. This makes Atlas cleaner and more reliable. Geldux Atlas is being shaped around real trading context. Not prediction. Not fake signal dashboards. Not random AI labels. The goal is to show useful market state around open interest, funding pressure, liquidation context, position health, freshness, and risk conditions before and after a user takes action. We also pushed the visual intelligence layer further. Atlas now has a more visual direction with perps heatmaps, liquidation pressure views, funding pressure views, Base protocol pressure, liquidity context, and shareable risk cards. The intention is to make risk easier to read at a glance without turning every color into a trade signal. A pressure map should help users understand conditions, not push them into action. Another area we improved is honesty around data. If something is missing, stale, not indexed yet, or unavailable, the interface should say that clearly. We are avoiding fake liquidation zones, fake volume, fake PnL, fake health scores, and fake intelligence. In a trading product, clear empty states are not a small UX detail. They are part of trust. Atlas Daily is also becoming a separate surface in the product direction. The full Geldux app remains the trading terminal. Atlas Daily is a lighter read only risk brief for people who want to follow market context without connecting a wallet or touching trading actions. No faucet, no contract writes, no position management. Just a cleaner public entry point into Geldux intelligence. We also cleaned the chart architecture. The trading chart now uses a first party API route. The browser calls the Geldux chart endpoint, while the server handles candle fetching and normalization privately. This keeps implementation details out of the frontend and gives us a cleaner path to change data sources later without changing the UI. The flows we care about now are: faucet cross margin isolated margin partial close full close portfolio history points Atlas mobile UX Geldux is still early, but the shape is becoming clearer. Execution is one part of the product. Risk context is the layer that makes the experience more useful. That is where the work is going.

Days like this are exactly why Geldux is being built. When BTC is down hard and the whole market feels heavy, trading becomes less about confidence and more about risk control. The questions change fast: Am I overexposed? Is leverage crowded? Are liquidations getting closer? Is funding turning one sided? Is the data fresh enough to trust? Should I trade, reduce, wait, or do nothing? Most trading products make it easy to click long or short. But on days like this, traders need more than execution. They need market context, position health, liquidation awareness, funding pressure, and honest data freshness before taking risk. Geldux is still testnet-stage. But the mission becomes clearer every time the market bleeds: build a trading terminal that helps traders read risk before they act. Trade with a map, not blind candles. #crypto #btc #base

Onchain trading should not feel like stopping at a wallet popup every few seconds. That is why we are testing session trading on Geldux. The idea is simple. A user approves a limited trading session once, then trades faster inside that approved scope. This does not mean unlimited wallet access. It does not mean the app can do anything with the wallet. A session should be controlled by clear permission limits. For example: which market can be traded which actions are allowed maximum collateral size maximum leverage maximum trade size allowed margin mode maximum slippage acceptable price range session expiry time oracle freshness requirement risk checks before execution So instead of giving broad approval, the user gives a specific permission for a specific trading window. That is the main difference. A normal approval can feel open ended. A session should feel limited, temporary, and purpose built for trading. In practice, this can make the trading flow much smoother. Open a position. Increase a position. Reduce a position. Close a position. The user should not need to break focus and sign again for every small step, as long as the action stays inside the session limits. But speed alone is not the goal. Security matters more. A Geldux trading session should not be able to withdraw funds freely. It should not be able to change user settings outside the session. It should not be able to execute unrelated wallet actions. It should not be able to trade markets the user did not allow. It should not be able to exceed the user’s approved collateral or leverage limits. It should not keep working forever after the session expires. Every session should have boundaries. Every trade should still respect risk controls. Every permission should be understandable before the user approves it. This is the experience we are building toward: fewer interruptions faster trade execution clear permission limits self custody onchain settlement risk checks still active Session trading is not about removing user control. It is about reducing unnecessary friction while keeping user control clear. This also connects with Geldux Atlas. Before trading, users should be able to see risk context like market pressure, funding pressure, long and short imbalance, liquidation pressure, position health, and freshness. So the full Geldux direction is not just faster execution. It is faster execution with better context and safer permission boundaries. Trade with fewer interruptions. Trade with clearer limits. Trade with a map, not blind candles.

Hey WAY Too Based, happy to be here. We’re Geldux, a Base-native perps trading and risk-intelligence product currently on testnet. The idea behind Geldux is simple: onchain trading should not only be about clicking long or short. Traders need better context before taking risk. Geldux is being built around two connected layers: a perp trading terminal for cross and isolated margin and Geldux Atlas, a risk-intelligence layer for market pressure, open interest context, funding pressure, liquidation awareness, position health and No Trade Radar. Still early. Still testnet. But the direction is clear: execution should not live alone. A better Base trading product should make risk easier to read before a trade is opened, increased, reduced or closed. Excited to connect with more Base builders and traders here. Feedback is always welcome. Trade with a map, not blind candles.

Builder note #5 from Geldux: We have been working deeper on Geldux Perps V2. Not only on the interface. On the system underneath it. Earlier, the product was more focused on making the trading flow visible and usable. Now the work is about making the whole perp stack cleaner: execution collateral accounting oracle safety margin logic position lifecycle risk reads market intelligence The biggest change is structure. Trading actions now sit on top of clearer internal layers instead of everything feeling like one combined flow. Collateral and liquidity accounting are handled more carefully, including reserved balances, fee buckets, insurance balance and bad debt tracking. Price validation is tighter through fresh oracle checks, confidence limits and normalized mark prices. The trading engine supports the full position lifecycle: open increase partial close full close cross collateral deposit cross collateral withdrawal isolated margin cross margin long and short exposure funding updates open interest accounting liquidation handling fee routing acceptable price protection The read side is cleaner too. The frontend can now pull better previews, account risk, position health and liquidation related data without forcing the trading UI to guess too much. That matters for Atlas. Risk intelligence is only useful when protocol state is structured properly. Position Health becomes clearer. Liquidation Radar becomes more grounded. No Trade Radar becomes more explainable. Funding pressure and OI context become easier to read. The current setup is intentionally controlled: 5 curated perp markets 20x max leverage 10 gUSD minimum collateral Pyth based price validation market level risk settings cross and isolated margin support Still testnet-stage. But the product is becoming stronger in the right way: less fragile execution cleaner accounting better risk visibility more useful intelligence before a trade is taken Geldux is not just trying to make perps clickable. We are building toward a trading terminal where execution and risk context work together. Trade with a map, not blind candles.

Builder note #4 from Geldux: The Intelligence layer is where Geldux starts to feel like more than a trading screen. Before opening a perp position, traders should be able to understand the market around them. Not just price. But pressure. Who is leaning long? Who is leaning short? How much open interest is active? Are tracked positions mostly winning or losing? Is the setup crowded? Is my own position sitting in a risky place? That is the goal behind Geldux Intelligence. It gives traders a cleaner view of market pressure and personal exposure before more risk is taken. Still testnet. Still early. But the direction is clear: Trading should come with context. Perps should not feel blind. Trade with a map, not blind candles.

A perp terminal should help answer more than: “Do I long or short?” Before taking risk, traders should be able to ask better questions: Is this market crowded? Is leverage building too fast? Is funding becoming one-sided? Is liquidity thin? Are liquidation zones nearby? Is the data fresh enough to trust? Is doing nothing the better trade? That is the direction behind Geldux Atlas. We are still testnet-stage, but the product principle is simple: Execution should not live alone. Risk, market context, position health, and data freshness should be part of the trading experience. Perps should not feel blind. Trade with a map, not blind candles. #base #prep #intelligence

Weekend builder note from Geldux: Some updates are not loud. No new flashy screen today. Just cleaner infrastructure behind the trading experience. We tightened the chart data path so the trading UI can stay simple: open market switch timeframe read candles check volume continue trading Behind that, the app should handle candle fetching, normalization, fallbacks, and source details cleanly. That matters because trading products should not feel fragile. If candles are available, show them clearly. If data fails, fail cleanly. If the source changes later, the frontend should not need to be rebuilt around it. Still testnet-stage, but this is the kind of weekend work that makes Geldux stronger: cleaner data paths better chart reliability less client-side mess no fake candles no noisy broken states Small plumbing. Better product foundation. Trade with a map, not blind candles. #base @buildonbase