Please report vulnerabilities privately to m@b402.ai. We'll acknowledge within 72 hours and aim to ship a fix within 14 days for high-severity issues. Coordinated disclosure timeline is 90 days from acknowledgement; we'll publish details after a fix is in production.

Do not file public GitHub issues for security reports.

In scope:

• The bot's webhook handler (src/copy-trade/) — anything that lets a third party trigger spurious copies, drain a user's balance, or read data they shouldn't.
• The setup wizard (src/setup/) — anything that leaks the bot token, Helius key, or follower wallet via the generated .env or its logs.
• The SQL schema (sql/) — privilege-escalation or data-leak issues.

Out of scope:

• The b402 shielded pool itself. Report there: https://github.com/mmchougule/b402-solana/blob/main/SECURITY.md
• Solana validator / RPC bugs.
• Telegram protocol issues.
• Issues that require an attacker to already control a follower's bot token or Postgres credentials.

The bot's threat model and trust assumptions are:

• The hosted b402 relayer is trusted to not censor or front-run, but cannot spend user notes (it never sees spending keys). Anyone can run their own relayer.
• The Helius RPC is trusted for tx data freshness. A malicious Helius could withhold leader events or feed forgeries; the latter is caught because the bot re-derives swap amounts from on-chain transfer arrays.
• The Postgres database is trusted to not be tampered with by an external attacker.
• The Telegram channel is trusted as the authenticated user surface; bot tokens leaking compromises the per-bot deployment.

For the full b402 protocol trust model see https://docs.b402.ai/solana/concepts/trust-assumptions.