ErdChainTrace started with a simple problem, i wanted to follow the money on MultiversX but I couldn’t find a tool that showed the flows clearly. The first version used a tree layout around the address being traced, followed by a force simulation to arrange the graph it worked for small traces, but it quickly became difficult to read as more addresses were added. You could see the connections, but not clearly where the money was going. I also realised that tweaking the simulation forces barely changed the result the initial layout was doing most of the work. So I decided to build my own rendering engine from scratch. I called it Ariane. Instead of relying on physics simulations, Ariane uses the Sugiyama method to arrange addresses in layers. The address being investigated stays in the middle, incoming flows are placed on the left and outgoing flows on the right. Loops are displayed as return arcs, and the order of addresses in each column is calculated to reduce crossing lines. The whole layout is calculated in under 100 ms for graphs with 700 bubbles. More importantly, expanding an address doesn’t move everything else around. If you move a bubble manually, it stays where you put it. Labels also have their own space, so they don’t end up on top of each other. Large traces needed another solution. Showing every single address isn’t always useful, especially when a few accounts handle most of the value. Ariane can group smaller addresses while keeping the most significant ones visible. On one Binance trace, that reduced the graph from 753 bubbles to 60. I’ve been testing the changes against real traces. On the same example, Ariane reduced the number of line crossings from 129 to 12 compared with the old Network mode, which I’ve now removed. I’m currently working on V2 the main addition is Atlas, a database of more than 2200 identified services across exchanges, DEXs, bridges, protocols and staking providers. I’m being careful with labels. An address is marked as confirmed only when there’s a reliable basis for linking it to a service, such as an address published by the project or on-chain evidence linking a contract to it. Otherwise it’s marked as probable. Individual wallets won’t be labelled as belonging to specific people. There’s still work to do on the investigation interface and on exports that make it easier for anyone to check the findings directly on the chain. V2 is still in development. I’d rather take the time to test everything properly than release it too early. In the meantime ErdChainTrace remains free and open to everyone. And as always, feedback is welcome.