ProofKit is an open network: anyone can run a node, and operators are not identified. Safety has to come from how routes and checks work, not from trusting whoever runs a node. Here is what protects you and where it stops.
1 · What one node can see
On a direct route the exit node sees your IP address and the sites you connect to, like any VPN server. On a 2-hop route the entry sees your IP address but not your destinations, and the exit sees your destinations but only the entry's address.
Limit: that split helps only when the entry and the exit are run by different operators. Nothing stops one operator from running both, and the marketplace does not show who runs a node. Someone who controls both ends, or can watch both, can match them up.
2 · What is logged
ProofKit's servers keep your account, your routes and their traffic totals, and no IP addresses; the node software keeps logging off for Xray and the entry proxy. The details, including what the olcRTC server logs, are in What is logged.
3 · Checks against billing fraud
- Probes: ProofKit's servers download test files through each exit and compare the bytes with what the node reports.
- Cross-checks and anomaly detection: entry-versus-exit byte comparisons, network-card counters and statistical checks raise flags. See How nodes are checked.
- Stake: each node locks 1 TON; 3 or more critical flags in 24 hours cost it 10% of its stake. See Stake.
- Checkpoints: nodes sign their traffic counters every 30 minutes; see Traffic checkpoints.
4 · A software self-check
A config-verified badge means the node's software reported that its logging is off. It deters casual tampering; it is not a guarantee that nobody captures traffic.
What you can do
- Use 2-hop when you would rather the exit did not see your IP address.
- Prefer nodes with a good health score and uptime; see Node health score.
- Treat a VPN as one layer. No VPN, ProofKit included, makes you anonymous on its own. For strong anonymity, use Tor.