Onchain Pipeline
The next incident looks like the last one.
We turn every on-chain incident into a detection rule and run it across the networks, including protocols nobody asked us to check.
Not in source code. In behavior.
01
An incident becomes a rule.
We take the root cause and describe how it looks in normal operation. Rules stay free of incident specifics (addresses, functions, names), so they match the whole class and not only the case they came from.
02
It looks normal. That’s where the signal is.
Routine activity already carries information about a weakness. The rule reads what protocols actually do.
03
Run on everything.
Every network we cover, all available data and patterns. No client list, no target list. Unaudited protocols included.
04
Uncertainty drops at every step.
Checks run in order of reliability, cost and time. What a deterministic check can settle, it settles first. Cheap models go before expensive ones. Fork replay has the last word.
05
Every incident widens the scan.
New incident, new rule. The library only grows.
About