$AI

reading this pool's swap logs…

• · We could not read $AI's metadata getters just now. So nothing is shown above rather than an empty list — an empty list would read as “this token publishes nothing”, which is a claim about the token made out of our own failed read.
• · Read them yourself: the contract on the explorer ↗

• · V4 depth is UNMEASURED, not zero: V4's PoolManager is a singleton holding every pool's tokens in one contract balance, so per-pool reserves cannot be read as balances. We say so rather than guess.
• · Volume is the 24h ETH leg the chainwide index attributed to each pool — a floor over what its per-tick budget could walk, never a lifetime total. Share divides by $AI's 24h ETH across all 1 of its pools, not just the rows listed.

last 24h 0 ETH added over 0 events · 0 ETH left the pools over 0 movements (V3 collects and V2 burns — the two ways tokens actually leave a pool contract) · 2 V4 events carry a liquidity delta and no amounts, so they are counted here and add nothing to the ETH figures. Floors over the pools and blocks we walked, not a census.

• Every row is a log we decoded, stamped with the block it came from. This panel states what moved, when, and who signed for it — it does not label a removal, and it never will. Taking liquidity out is an ordinary operation; the reader draws the conclusion from the numbers and the times.
• V3 splits a removal in two: closed is the burn, which only credits what the position is owed, and withdrawn is the collect, where the tokens actually leave the pool. A collect carries accrued fees as well as principal and the event does not separate them, so we do not either. V2 does not split: a V2 removed transferred in the same call, with no collect to wait for, which is why the summary counts it as ETH that left rather than as principal credited.
• V4 rows report a liquidity delta (Δ) because that is all the event says. Turning one into ETH needs the pool’s price at that block, which is not in the event and not in the receipt — V4’s ETH side is native and moves with no log. Their ETH column reads “unmeasurable”, never 0.
• A signer shown as unread is one we did not resolve inside this tick’s block-fetch budget. It is never backfilled from the event’s own actor, which on this chain is a position manager essentially always.
• Percentages are deliberately absent. “80% of the liquidity left” needs the depth before the removal, and we hold depth only as a live read at one block — no history to divide by. The absolute amounts above are what we actually measured.

• · Every row is the account that SIGNED the swap transaction — never the router the trade was routed through. (Measured while building this: every swap-event “sender” in a live window was a contract; a board of those would be a list of routers.)
• · Totals cover ONLY the indexed window: 145 swaps attributed for $AI between 2026-08-07 10:31Z and 2026-08-07 10:37Z — at least 75% of the 192 swaps our volume index counted over the same hours. Attribution is budgeted per tick and takes the highest-value blocks first, so every figure here is a floor over a stated period, never a lifetime total.
• · ETH in and ETH out are shown separately because net alone cannot tell “never sold” from “round-tripped”.
• · An account signing on others' behalf (a keeper, a relayer) appears as one signer. The “tokens” column is shown so bot- and infrastructure-shaped accounts are visible rather than silently ranked as people.

We could not measure holder distribution for $AI: our node did not answer, so no holder read was attempted. Nothing is claimed about this token's distribution. No percentage is shown, because a failed read of ours must never render as a number about someone else's token.