Do not open a public GitHub issue for security vulnerabilities.

Report privately via GitHub Security Advisories:

1. Go to https://github.com/Lians-ai/Lians/security/advisories
2. Click "Report a vulnerability"
3. Fill in the details

Alternatively email security@lians.dev with:

• Description of the vulnerability
• Steps to reproduce
• Potential impact
• Any suggested fix

We will acknowledge receipt within 48 hours and provide a remediation timeline within 7 days.

In scope:

• SQL injection or data isolation bypass in the API layer
• Authentication/authorisation bypass (X-API-Key, X-Admin-Secret)
• Information barrier (RLS) bypass allowing cross-namespace data access
• Cryptographic weaknesses in the AES-256-GCM per-subject encryption
• Audit chain (SHA-256 hash chain) tampering or bypass
• Remote code execution via any input path

Out of scope:

• Issues in dependencies that are already publicly disclosed upstream
• Theoretical attacks with no practical exploit path
• Denial-of-service via resource exhaustion without auth bypass

Lians enforces isolation at multiple layers:

• API key authentication — all data-plane routes require X-API-Key; keys are stored as SHA-256 HMAC hashes, never plaintext
• Admin secret — a separate credential (X-Admin-Secret) gates all /v1/admin/* routes; it is not derivable from agent keys
• PostgreSQL RLS — information barriers are enforced at the database layer with FORCE ROW LEVEL SECURITY, not the application layer
• AES-256-GCM encryption — each subject's memories are encrypted under a unique data-encryption key (DEK); destroying the DEK makes the data cryptographically unrecoverable (GDPR crypto-shred)
• Append-only audit chain — every write is recorded in a SHA-256 hash chain; audit rows are never updated or deleted

We follow coordinated disclosure. Once a fix is ready we will:

1. Release a patched version
2. Publish a GitHub Security Advisory
3. Credit the reporter (unless they prefer anonymity)