Hey all,
The cell model is the single hardest thing about CKB to hold in your head. A raw transaction from a node is a wall of hex: capacities in shannons, code hashes, packed since values, molecule witnesses. Nothing about it tells you “Alice sent 100 CKB and 1,500 USDI to Bob and got change back.”
So I built ckb-viz to close that gap.
Live: https://ckb-viz.truthixify.dev
Code: Paste a transaction hash (or just open it, it loads the network’s latest tx) and
you get:
- The cell flow: inputs → the transaction → outputs, connected, with capacity and the fee visible as the difference between the two sides.
- Decoded scripts: known locks/types named (Secp256k1, Omnilock, JoyID, xUDT, Nervos DAO, Spore…), unknown ones clearly marked, never guessed.
- A plain-language summary: “A token transfer of 19,310.74 USDI”, “A Nervos DAO withdrawal” (with the real fee split out from the interest earned), and so on.
- Cell lineage: trace an input back to where it came from, an output forward to where it was spent.
- Full cell detail: ckb2021 address, code hash, args, raw and decoded data, witnesses, and inline Spore image previews.
It reads directly from a public node (no backend), works on mainnet and testnet, and every transaction is a shareable link.
Where this is going: the honest aim is something like Tenderly, but for CKB, the kind of dev/debug tooling EVM has and we don’t. This visualizer is the starting point (the “transaction overview” layer). Next on my mind: why did a tx fail (and which script), then dry-run/simulation, and eventually a CKB-VM debugger with cycle profiling.
It’s early, and I’d genuinely love feedback on what’s confusing, what’s missing, or what you’d want next. And if you find a transaction it decodes wrong or doesn’t recognize, send me the hash. 