Public financial evidence: Tornado root-window and Car comparisons

Prepared on 1 November 2026. Read REPORT-DATA.json for the reviewed findings.

Current state and recent events: current/CURRENT-FINDINGS.json
Withdrawal signatures, frames and root window: roots/ROOT-WINDOW-FINDINGS.json
Exact financial comparisons: joins/FINAL-SUMMARY.json

Offline replay from the extracted collection:
  python3 root/replay_offline.py current/verify_current_offline.py
  python3 root/replay_offline.py roots/verify_root_window_offline.py
  python3 root/replay_offline.py joins/verify_joins_offline.py

The replay wrapper denies network and child processes. Read the explicit limits.
They replay provider-returned records, signatures and consistency checks;
they are not a standalone consensus proof, Solidity compiler or Groth16 verifier.

The 111 withdrawals are pool context, not 111 attacker transactions.
No withdrawal is attributed to C91 and no Car financial link is established.
Disclosed accounts, controller keys and trading counterparties are separate sets.
The 0.0011 pUSD bulk receipt does not establish action by its recipient.

Full third-party source and social-post bodies remain local. Public projections
contain technical metadata, source links/hashes, financial records and own code.
The official source-capture script is online collection code, not an offline verifier.
Older financial/attribution observations keep their own source and scope limits.
