Frequently Asked Questions

Sentinel is designed with player safety as a top priority. Our SDK operates entirely in user-mode — we don't install kernel drivers or make low-level system modifications that could destabilize systems.

What we don't do:

• Install kernel-level drivers
• Access personal files or browser data
• Scan processes outside your game
• Persist after your game closes

Our approach focuses on protecting the game's memory space and detecting tampering within that scope. This is less invasive than some anti-cheat solutions while still providing strong protection.

Hot-patching allows us to push new detection rules to your game without requiring a client update. Here's how it works:

1. Detection: When we identify a new cheat, our research team analyzes it and creates a detection rule.
2. Deployment: The rule is deployed to our global edge network.
3. Application: Your game's SDK checks for updates periodically (configurable, default every 5 minutes) and downloads new rules.
4. Activation: Rules are applied in-memory without restarting the game.

This means we can respond to new cheats in minutes rather than waiting for your next game update cycle. For more technical details, see our blog post on live patching.

We collect the minimum data necessary for anti-cheat functionality:

• Hardware identifiers: Used for ban enforcement and preventing banned users from creating new accounts
• Game session metadata: Session ID, timestamps, game version
• Detection events: When suspicious activity is detected

What we don't collect:

• Personal information (name, email, etc.)
• Browsing history or cookies
• Information from other applications
• Screenshots or screen recordings
• Keystroke logs

All data is encrypted in transit and at rest. See our Privacy Policy for complete details.

Sentinel is designed for performance-critical games:

• CPU overhead: Less than 1% in typical usage
• Memory usage: Under 20MB
• Network bandwidth: Negligible (a few KB per session)
• Startup time: Under 100ms initialization

Our SDK is written in optimized C++ and runs checks asynchronously to avoid blocking your game's main loop. Intensive scans are distributed across frames to prevent stuttering.

We test on minimum-spec hardware to ensure our SDK doesn't create issues for players with older systems.

Currently supported:

• Unreal Engine 4.27+ and 5.x
• Unity 2020.3+ (IL2CPP and Mono)
• Custom C++ engines (Windows)
• Roblox (via Watchtower)

Platform support:

• Windows 10/11 (64-bit) — Full support
• Windows 7/8 — Limited support
• macOS — In development
• Linux — In development

For other engines or platforms, contact us to discuss your requirements.

You have full control over how detections are handled. Options include:

• Shadow mode: Log the detection but take no action (useful for testing)
• Soft ban: Move the player to a separate matchmaking pool
• Kick: Disconnect the player from the current session
• Ban: Permanently prevent the player from accessing the game
• Custom callback: Trigger your own logic

We recommend starting in shadow mode to evaluate false positive rates before enabling enforcement.

False positives are our top priority to avoid. Our approach includes:

• Conservative detection: We only flag behavior we're highly confident is cheating
• Multi-signal verification: Multiple checks must agree before taking action
• Confidence scoring: Low-confidence detections are logged, not enforced
• Extensive testing: New rules are tested against millions of legitimate sessions

If a false positive does occur, we provide tools for you to review and reverse bans. Enterprise customers also have access to our support team for escalation.

Yes, Sentinel is designed to work alongside other security tools. We don't interfere with:

• Platform-level anti-cheat (Steam, EAC, BattlEye)
• DRM solutions
• Crash reporting tools
• Analytics SDKs

Some customers use Sentinel as a secondary layer of protection, especially for detecting game-specific exploits that generic anti-cheat solutions might miss.