Human rulings assign explicit 'correct' or 'incorrect' verdicts that lock the claim against automated overwrite while recording every change in an append-only audit log.

flowchart LR
    A[Human Review] -->|Rule Correct / Incorrect| B[Lock Claim: is_human=True]
    A -->|Withdraw Rule| C[Reset Claim: is_human=False]
    B --> D[Append to claim_rulings]
    C --> D
    B --> E[Prevent Checker Overwrite]
    C --> F[Return to Research Engine]