Skip to main content

​
Endpoint

POST /v1/trees/{tree_index}/proofs Returns multiple Merkle inclusion proofs for the specified insertion indices, all guaranteed to be computed against the same tree root. This is what the burner factory calls per batch — burns that bundle multiple Stealth Pool Notes need all proofs to share a root.

​
Path Parameters

  • tree_index (u64) — Zero-based index of the Merkle tree.

​
Request Body

JSON object with a single field:
  • insertion_indices (array of u64) — Leaf positions to generate proofs for. Maximum 8 indices per request.
Example: 
{
  "insertion_indices": [42, 108, 255]
}

​
Response

Protobuf BatchProofResponse containing an array of proof objects. Each proof has the same structure as the single proof endpoint:
  • root (string) — 64-character little-endian hex-encoded Poseidon root hash. Identical across all proofs in the batch.
  • tree_index (int64) — Echo of the path parameter.
  • insertion_index (int64) — The leaf position this proof corresponds to.
  • proof (array of 20 strings) — Sibling hashes from leaf to root, each 64-character hex.
  • leaf (string) — 64-character hex-encoded final commitment.

​
Why Batch?

Single proof requests (GET /v1/trees/{tree_index}/proof/{insertion_index}) do not guarantee a consistent root across multiple calls — the tree may grow between requests. The batch endpoint acquires a read lock on the tree, ensuring all returned proofs share the same root. This is required for Groth16 ZK proofs that reference multiple note commitments.
The SDK’s burner factories call this endpoint internally per batch. Receiver-burnable → ETA burns chunk to ≤4 notes per proof; self-burnable burners loop one note at a time.