midnight-proofs:proof-caching

You can cache ZK proofs by deriving keys from circuit and witness data, storing verification results to avoid redundant work, and applying TTL-based invalidation to maintain proof freshness.

Proof cache invalidation triggers automatically through TTL-based expiration and state-change-aware purges, removing stale proofs when underlying circuit inputs or contract state changes occur.

Yes, Redis-backed distributed caching is supported for multi-instance deployments, enabling shared proof storage and batch operations across high-volume verification environments.

Batch operations allow simultaneous proof lookups and sets, maximizing throughput by processing multiple cached proofs at once in high-volume verification and generation environments.

ZK proof caching is not ideal for circuits with unstable inputs or rapidly changing states, as frequent invalidation negates performance gains and increases computational overhead.