indexCross-Chain Messaging & Bridges#cross-chain#bridges#messaging

Cross-Chain Messaging and Bridges

Introduction

Cross-chain systems create claims in one execution domain based on events or state from another. The branch studies message verification and lifecycle before token movement abstractions.

Why It Matters

Bridges combine high-value accounting with mismatched finality, validator/relayer assumptions, asynchronous retries, wrapped assets, upgrade keys, and operational recovery. Collapsing those states creates replay, double-mint, stuck-message, and insolvency risk.

Mental Model

A message binds source/destination chain and contract, payload, sender, nonce/order, source finality evidence, verifier, execution status, retry policy, and accounting effect. Pending, accepted, executed, failed, and recovered are distinct states.

Questions This Branch Answers

  • Who or what proves a source event to the destination?
  • How do replay, ordering, reorg, retry, and partial failure interact?
  • Which controls limit damage when verification or liquidity fails?

Scope

Bridge taxonomy, lock/mint/burn/liquidity models, cross-domain messages/nonces, replay/idempotency/ordering, finality mismatch/reorg recovery, rate limits/pauses/accounting, canonical bridges/CCIP/IBC, and incident lessons.

Out of Scope

Live asset bridging, bridge yield, universal interoperability claims, production validator sets, and ranking protocols without versioned architecture and evidence.

Dependencies

finality/reorgs, L2 settlement, token behavior, and distributed messaging.

Candidate Note Roadmap

  • bridge-taxonomy-and-trust-models — Classify verification, custody, liquidity, admin, and recovery assumptions.
  • lock-mint-burn-and-liquidity-networks — Reconcile canonical and synthetic asset movement models.
  • cross-domain-messages-and-nonces — Bind domains, identities, payloads, and message uniqueness.
  • replay-idempotency-and-ordering — Make retries safe under duplicate and reordered delivery.
  • finality-mismatch-reorgs-and-recovery — Simulate source rollback after relay or destination execution.
  • rate-limits-pauses-and-accounting — Bound exposure and retain reconciliation during incidents.
  • canonical-bridges-ccip-and-ibc — Compare concrete verification and channel architectures.
  • bridge-incidents-and-wrapped-asset-risk — Reconstruct primary failure mechanics and residual claims.

Future Project

A two-domain Cross-Chain State Mirror with controllable finality, relayer retries, at-most-once execution, explicit message states, per-chain accounting, rate limits, and reorg/compromise incident drills.

Initial Invariants

Each valid message executes at most once; retries are idempotent; source/destination/domain fields cannot be substituted; minted/locked/burned/released accounting reconciles; pending, failed, and finalized states never collapse.

Initial Threat Model

Verifier/validator/admin compromise, source reorg, forged or replayed messages, inconsistent ordering, liquidity exhaustion, wrapped-asset depeg, pause abuse, rate-limit bypass, upgrade takeover, relayer censorship, and recovery double-spend.

Primary Sources

Canonical L2 bridge specifications, Chainlink CCIP and Cosmos IBC specifications, protocol postmortems, and sources catalogued in SOURCES.md.

Connects to: rollups, privileged controls, and reconciliation/response.