Every AI decision you seal with sebbi.pro keeps its inputs and its verdict in the chain, and the engine gives the same answer to the same inputs every time. So you can take a new rule, a tighter threshold or tomorrow's law, and replay every real decision you have ever made under it. In seconds you see which verdicts would have changed, which way, by month and by money, and the run itself is sealed into Bitcoin as evidence.
Move the thresholds, weigh factors up or down, add high-risk countries, re-weight your own risk signals, or add hard rules on any field you send.
Every sealed decision in your chosen period is checked untouched, then replayed. Only decisions the current rules reproduce exactly are counted.
Exactly which verdicts would flip, which way, month by month, the money behind each change, and the rule that caused it.
The rule set and the results are fingerprinted and sealed into the chain. Share the report with an auditor, an insurer or a regulator.
Most AI compliance tools keep records you only open after something goes wrong, and their scoring runs through a model, so a decision can never be performed twice. sebbi.pro's scoring is arithmetic. Your history becomes a test bench: tune your guardrails against what really happened, before a new penalty ever applies.
No. Each record's stored hash is recomputed first, then today's rules are replayed to prove they give back the original verdict. Anything that doesn't reproduce exactly is listed separately, never counted.
Never. Your sealed history is only read. The replay run is added to the chain as a new record of its own.
Only people you give the report link to. Use it as evidence for an auditor, insurer or regulator.
It's included with every sebbi.pro key: 7 days free, then 50p per device per month. Checking any proof is always free.
AI agents can run it too: sebbi_forensic_replay on the MCP server at https://sebbi.pro/mcp.