Reputation scoring in RubyScore is a set of methods and models that turn on-chain activity into reputation signals and scores. Based on these signals, three independent products are formed:
A unified score that aggregates a wallet’s behavior across multiple networks. To increase MRS, a wallet should demonstrate regular and diverse activity across different chains (not just one-off actions for a single drop).
A local proof of “human-ness” based on on-chain actions within a specific ecosystem (network or project). For each ecosystem, a custom formula (metric weights and thresholds) is configured on request to determine the PoH ID levels.
Achievement levels for a wallet based on on-chain activity within a specific ecosystem. Level progression depends on sustained engagement over time: the higher the level, the more active and consistent the behavior in that network/project.
Amount on balance
Total gas spent
Transactions with unique contracts
Transactions on different days
Transactions on different weeks
Transactions on different months
Transaction volume
Number of transactions
Additionally considered: diversity of interaction categories (DeFi/NFT/Bridge/DAO), depth of interaction sequences (e.g., bridge → deposit → swap → LP → vote), temporal stability, economic rationality, and anti-manipulation signals (filtering scripted and repetitive patterns).
Last updated
