Retraces every reallocation ever applied to a campaign, from its latest merkle tree — the newest root whose tree the tree service holds, named in atRoot. The distribution chain is inferred from the campaign ID.
A reallocation is a two-step protocol recorded in the tree, and it never changes a campaign's total. Step 1 shrinks the original recipient's leaves to what it had already claimed and stages a marker at the zero address (tmp_reallocation_{from}_{to}). Step 2 drains that marker and either credits the destination (reallocation for a creator request, unclaimedRewards for a wipe sweep) or returns the reward to the original recipient (fail_reallocation / fail_wipe) when it claimed in between.
Drained markers survive at amount zero, so the senders roster and the inFlight, returned and per-destination settled amounts are exact.
Not available here: per-sender settled and stagedAt (walk is null). A destination's leaf aggregates every sender and a drained marker has no amount left, so recovering those means one tree read per root — see POST /v4/reallocations/report/{campaignId}/walk.
Authentication: Requires either a valid x-api-key header or the back-office bearer secret.
⚠️ Beta: this endpoint is in beta. The behavior, response shape and data format may change at any time without notice.
## Path parameters - `campaignId` string, required: **On-chain campaign ID** (32-byte hex hash, e.g. `0x93cf385c...`). The distribution chain is inferred from it. NOT internal `Campaign.id`. See [ID types](/integrate-merkl/id-types). ## Response `200` Response for status 200 - object - `campaignId` string, required: Onchain campaign ID the report covers - `distributionChainId` number, required: Distribution chain inferred from the campaign ID - `rewardToken` object, required - `address` string, required - `symbol` string, required - `decimals` number, required - `atRoot` object, required - `root` string, required: Merkle root the report describes — the chain's newest one the tree service can read. - `timestamp` string, required: Unix timestamp (seconds) of that root - `totals` object, required - `inFlight` string, required: Sum of every marker still staged, in raw token units - `settled` string, required: Sum held at every destination under a reallocation reason, in raw token units - `returned` string, required: Sum returned to senders by a failed settle, in raw token units - `staged` string: Sum of every step-1 wave the walk observed, in raw token units. Only present on a walk-job result. - `destinations` object[], required - `to` string, required: Address credited with the reallocated rewards - `kind` union, required: `creator-request` — the creator staged an on-chain reallocation (`reason = reallocation`). `wipe` — the wipe job swept unclaimed rewards older than a year to the Merkl dumper (`reason = unclaimedRewards`). - string - string - `settled` string, required: Amount held at this address under the reallocation reason, in raw token units. Aggregated over every sender — per-sender attribution needs the walk job. - `senders` object[], required: Every original recipient ever reallocated for this campaign, sorted by amount moved descending. - `from` string, required: Original recipient the rewards were taken from - `to` string[], required: Destination(s) this sender's rewards were staged toward, parsed from its `tmp_reallocation_{from}_{to}` marker(s). More than one means the sender was reallocated in separate waves to different addresses. - `inFlight` string, required: Amount staged at step 1 but not yet settled, in raw token units. - `returned` string, required: Amount returned because the sender claimed between step 1 and step 2 (`fail_reallocation` / `fail_wipe`), in raw token units. The sender kept these rewards. - `settled` string: Amount taken from this sender that actually reached its destination, in raw token units. Only present on a walk-job result, and only when the walk covered every wave. - `stagedAt` object[]: Each step-1 wave that staged rewards from this sender, oldest first. Only present on a walk-job result. - `root` string, required: Merkle root of the tree the marker first carried this amount in - `timestamp` string, required: Unix timestamp (seconds) of that root - `amount` string, required: Amount staged in this wave, in raw token units - `to` string, required: Destination this wave was staged toward - `walk` object, required - `complete` boolean, required: `true` when every step-1 wave was accounted for, so each sender's `settled` is exact. `false` means the walk hit its snapshot budget, or the tree service no longer holds the trees the earliest waves are in, or markers disagreed with the staged total — and the amounts are lower bounds. - `incompleteReason` string: Why the walk could not close - `rootsScanned` number, required: Roots read — far fewer than the roots covered - `oldestRoot` string, required: Oldest root covered — the campaign's first, or the oldest whose tree the tree service still holds - `newestRoot` string, required: Newest root covered — the chain's latest readable one ## Changes - **2026-09-09** `020dd1a3d27d`: 1 info - endpoint added [Operation history](https://skmtc.dev/merkl/apis/merkl-api/changes/v4/reallocations/report/:campaignId/get.md) --- [API](https://skmtc.dev/merkl/apis/merkl-api.md) - [All operations](https://skmtc.dev/merkl/apis/merkl-api/llms.txt) - [OpenAPI document](https://skmtc.dev/merkl/apis/merkl-api/revisions/60ffe2c7d887?raw)