Hardening an Edge AI Platform: What a Real Security Release Looks Like (NeoMind v0.9.21)

Описание и применение

NeoMind v0.9.21 демонстрирует комплексную защиту: закрытие открытой регистрации, шифрование секретов, проверка путей, проверка пакетов, ревокация сессий, безопасные обновления, проверенные резервные копии и наблюдаемость.

Задача

Edge AI platforms expose a wide attack surface: open registration, plaintext secrets, path traversal, unverified packages, weak credential lifecycle, privilege escalation during upgrades, and inadequate data safety.

Результат

NeoMind v0.9.21 demonstrates a comprehensive security hardening process, addressing registration, secrets, path traversal, supply chain, credential lifecycle, upgrade privilege, data safety, and observability, resulting in a more secure edge AI platform.

Рабочий процесс

• Audit registration endpoints for open access
• Implement admin‑only registration with persisted setting
• Seal all secrets at rest using shared encryption key
• Add path‑join validation: reject dot‑segments, absolute paths, canonicalize and check containment
• Verify marketplace package checksums and enforce strict mode
• Revoke sessions on password change or user deletion
• Mask instance keys in API responses
• Stage upgrades in data/upgrade/ and trigger root apply via systemd path unit
• Add verified backup routine with manifest and retention
• Implement rollback guard by version‑stamping databases
• Expose Prometheus metrics for observability

Доказательства

POST /api/auth/register was a public, unconditionally‑open account‑creation endpoint

Источник

LLM provider API keys were stored as plaintext JSON in settings.redb

Источник

The API runs as the sandboxed neomind user (ProtectSystem=full + NoNewPrivileges=true)

Источник