Security
Security controls for a read-only crypto-risk intelligence service.
No custody
The service does not ask for seed phrases or private keys and does not sign user transactions. Token analysis is read-only. Payment uses a receiving address configured server-side; no wallet private key is stored by the application.
Server-side secrets
Database secret keys, RPC credentials, provider keys and dispatch secrets are server-only and must never use NEXT_PUBLIC environment variables. Logs exclude raw RPC URLs, payment credentials and report access tokens.
Worker isolation
Persistent jobs use database leases, fencing tokens, run isolation and atomic publication. A stale worker cannot publish or overwrite the result of a newer attempt.
Paid report access
Paid reports use a public report identifier plus a separate high-entropy access token. Only a SHA-256 hash of that token is stored. The token is carried in a URL fragment and Authorization header rather than a query parameter.
Responsible reporting
If you discover a security issue, do not include secrets, private keys or unnecessary personal data in the report. A dedicated security contact address should be added before public launch.