Skip to main content
Rhizome uses a hybrid data architecture:
This flow describes how data becomes available for querying and presentation. It does not make every layer equally authoritative.

​
Onchain protocol state

The contracts on Rootstock are the source of truth for current markets, positions, balances, oracle prices and freshness, and transaction results. Clients should use Rootstock RPC for safety-critical reads, transaction simulation, and submission.

​
Indexed and historical data

The indexer consumes Morpho events from Rootstock and builds queryable records for market discovery, account activity, liquidations, interest-accrual events, and transaction history. The current indexer uses Envio HyperIndex. Indexed data can lag, be unavailable, or temporarily differ during reorganization handling. It is a read model, not a transaction authority. A client must not use an indexed value to bypass a current onchain health check or simulation.

​
API and data layer

The API/data layer exposes indexed entities to applications and may combine them with direct RPC reads. It should preserve provenance: consumers should be able to distinguish historical indexed values from current onchain values. TODO: Publish supported public endpoint URLs, schemas, availability expectations, and versioning when they are finalized.

​
Frontend presentation

The frontend formats balances, rates, market metadata, and history for users. Presentation logic can estimate accrued values for display, but the frontend is not protocol state. Before submitting a transaction, it should refresh relevant RPC state and simulate the exact call.