State checkpoints

Almost all of XChain's history already lives on the blockchains it runs on. A small part (like the record of cross-chain trades) is stitched together by the validator network. State checkpoints capture that combined state, get signed by the validators, and are written back onto the chain itself. The result: anyone can prove the platform's state is honest, and any operator can rebuild it from scratch without asking a single peer for help.

The validators agree on a snapshot of each chain's state

↓

A 2f+1 quorum signs it, and one validator is elected to publish

↓

The signed checkpoint is written on-chain

↓

Anyone can verify state, or fully recover it, from the chain alone

What a checkpoint contains

Each checkpoint commits a fingerprint of a chain's state (its ledger, its actions, and its contract state) at a specific block. A fuller version also carries an archive of the cross-chain trade records, the one piece of platform state that isn't stored natively on any single chain, so nothing is left behind.

No single operator to trust

A checkpoint only counts if a two-thirds-plus (2f+1) quorum of staked validators independently signs it. The network then elects one validator to publish it on-chain and pay the fee, with a backup ladder if that publisher goes quiet. A dishonest validator can be outvoted, but cannot forge a checkpoint or sneak in a fake one.

Recover the whole platform from the chains alone

Because the checkpoints and the archived cross-chain records live on the blockchain, an operator who loses their database can rebuild complete, verified state from the beginning using only the coin chains. No asking other operators for a copy, no trusting someone else's backup. The chain is the source of truth, and the rebuilt state's fingerprints must match what was signed.

It can never touch your balances

A checkpoint is a read-only commitment. It never credits, debits, escrows, or moves a single token, so a missing or bad checkpoint cannot corrupt anyone's state. The worst case is simply that one fast recovery shortcut isn't available and the network falls back to a slower rebuild.

What this gives you

Independent proof that an explorer or validator is reporting honest state, fast disaster recovery for the people who run the network, and a platform whose entire history stays verifiable from public chains, with no trusted archive in the middle.