Sentinel Network is a backend API that gives game publishers a unified system for tracking device trust, managing bans across titles, and computing reputation scores — all through a single integration point.

Key Features

Device Verification

Cryptographic device identity using ES256 signatures with replay protection built in.

Multi-Scope Bans

Ban devices at the game, publisher, or global level. Supports cheat and social ban categories with optional expiry.

Reputation Scoring

Automatic reputation scores (0–100%) computed from ban history. Publishers control which scopes contribute to their scores.

Publisher Policies

Per-publisher enforcement policies that control which ban types and scopes are enforced and visible.

Publisher Dashboard

Web dashboard for managing publishers, games, API keys, bans, and team members with OAuth login and audit logging.

How It Works

  1. Devices are manufactured and registered by Sentinel — each device’s ES256 public key is provisioned at manufacture and stored on the server before the device ships.
  2. Game clients send signed payloads on each session start — the API verifies the signature, checks for replay attacks, and returns the device’s ban and reputation status.
  3. Publishers create bans when violations are detected. Bans can be scoped to a single game, all games by a publisher, or globally across the network.
  4. Reputation scores update automatically via a background worker that processes ban events and recomputes aggregates.

Architecture Overview

Client → APISIX + auth sidecar (API key auth) → Sentinel API → PostgreSQL
                                                             → Redis (cache + replay)
                                                             → Outbox → Worker → Reputation tables
APISIX and its auth sidecar validate API keys and inject trusted headers (X-Publisher-Id, X-Scopes) before proxying requests to the API. The client provides X-Game-Id directly.

Get Started

Follow the quickstart guide to make your first API calls in minutes.