GHO bridge activity

Message history checked Oct 10, 2026; analysis reviewed Oct 10, 2026. Chainlink’s CCIP API was paginated through the start of the window (9 pages). Current-day messages are excluded. This is a complete pagination of the provider’s indexed results, not an independent chain replay or guarantee against missing indexing.

564 of 564 messages in the overlapping pool-ledger period also match an indexed GHO transfer by transaction, sender and exact amount. The overlap runs Jul 12, 2026 through Sep 16, 2026. The download preserves those transfer log indices and block hashes. This cross-check uses an existing Blockscout ledger; it is not an independent replay of Ethereum.

The Aave address book identifies the Ethereum GHO token and pool. The cohort is all returned outbound messages carrying that token. It does not cover all Aave activity, all CCIP applications, or return bridges from other chains. Addresses use the API’s origin field. Bots, shared accounts, relayers and multiple addresses per entity can change the apparent repeat rate. Prior activity before this window is unknown, so the starting group is not a new-customer cohort.

Only the two exact governance launch message IDs are excluded from repeat-use calculations. Other operations may include treasury funding or rebalancing. No other traffic is labeled organic. Launch transaction evidence ↗

The inbound examples are separate from the outbound 90-day chart. Between October 8 and 11, five successful CCIP messages carried 6.130M GHO from Plasma (four messages) and Ink (one) to Ethereum. Two Plasma transfers went to the same address; three more reached three other addresses. Their receipts match the amount and recipient in Ethereum GHO-pool release logs. One Ink delivery took 22 minutes 38 seconds; the two smaller Plasma deliveries took 11 and 16 seconds. These are selected transfers from a bounded pool sample, not a route-wide count or latency distribution. Download the five matched examples ↗

Fee coverage is 70/842: two launch transactions and 68 selected messages. Examples are not a representative or random sample, so they cannot establish typical fees or a fee trend. Fee records retain their individual observation dates and exact raw amounts in the download. Fees are not extrapolated, converted using today’s prices or summed across currencies. See CCIP billing for the network/blockchain fee distinction. The separate inbound sample above reports four gross source charges in WXPL and one in WETH; these are not net LINK revenue.

Collection retains message identities and dated snapshots for later review. A new collection does not automatically replace this published analysis. Download the dated message dataset (JSON) ↓