Blockscout address metadata and indexed Robinhood Chain transaction history only.
Likelihood,
not AI identity.
RHC Lens estimates whether a wallet's execution pattern resembles manual or automated operation. It cannot distinguish an AI agent from a deterministic bot, script, or human using automation tools.
Every report exposes the strongest measured signals and their scores.
Fewer than 20 outgoing operations returns insufficient evidence.
How the score is assembled
No individual behavior proves automation. The heuristic score combines six supporting measurements.
TIMING REGULARITY
Measures interval consistency between successful outgoing operations.
REPEATED CALL SEQUENCES
Measures recurrence of adjacent target-and-method pairs rather than isolated duplicate calls.
DAILY ACTIVITY COVERAGE
Measures active-day density and average active hours without assuming an operator timezone.
RAPID EXECUTION
Measures the share of consecutive operations occurring within twenty seconds.
AUTOMATION INFRASTRUCTURE
Uses contract interaction and account-abstraction context without treating either as proof.
AMOUNT CONSISTENCY
Measures repeated non-zero transaction values as a weak supporting signal.
How to read the result
MANUAL PATTERN
Substantial timing and sequence variation consistent with manual operation.
MIXED SIGNALS
Evidence is unresolved; the wallet should not be described as partly proven.
AUTOMATION LIKELY
Multiple strong signals align with continuous or programmatic execution.
What can produce a false signal
- 01Professional human traders may use highly regular schedules and repeated strategies.
- 02Automated wallets may contain manual approval pauses or randomized execution timing.
- 03Shared wallets can combine several operators and automation systems in one history.
- 04Young wallets can change classification quickly as their sample grows.
- 05The score detects automation-like execution, not whether the automation uses artificial intelligence.
- 06The thresholds are engineering heuristics and have not yet been calibrated against a labeled ground-truth wallet dataset.