Skip to content

Roadmap ​

SentinelRN follows a phased plan. Every feature must map back to a documented threat.

Phase 1 — Trust core ✅ (0.1) ​

Risk engine, threat engine, AI guard (secrets, PII, prompt injection), redaction, policy engine, and the cohesive SentinelRN API.

Phase 2 — Platform & React ✅ (initial) ​

Android + iOS integrity detectors, pluggable provider with graceful fallback, React hooks + provider, new-architecture TurboModule codegen spec, and an example app. Ongoing: broader detector coverage and this documentation site.

Phase 3 — Attestation & advanced detection ​

  • Frida / runtime-hooking hardening
  • App signature / tampering validation
  • Play Integrity API
  • Apple App Attest + DeviceCheck
  • SSL pinning helpers
  • Secure logging helpers

Phase 4 — Ecosystem ​

  • Remote risk telemetry (opt-in)
  • OWASP MASVS mappings
  • Expo config plugin
  • Enterprise policy packs

The authoritative roadmap lives in ROADMAP.md.

Released under the MIT License. SentinelRN is risk-based — always enforce trust on the server too.