security-audit

To audit game codebase security vulnerabilities, a dedicated subagent inspects source files, assets, and configuration across six categories—save, network, input, data exposure, cheat, and dependencies—to classify findings by severity and produce a prioritized remediation report.

Before launching multiplayer game features, you should check for network exploits, save-file tampering, input validation gaps, data exposure, cheat vectors, and vulnerable dependencies, classifying findings by severity to prioritize remediation before public release.

Yes, you can run a security audit focused on network or save file scopes by parsing arguments to select full, network, save, input, or quick scopes, allowing targeted evaluation of specific security categories during pre-release or multiplayer feature development.

The security audit requires engine and language detection from technical preferences to evaluate save, network, input, data exposure, cheat, and dependency categories using a dedicated security-engineer subagent within the repository root.

The best way to block cheat vectors and input validation gaps is to spawn a dedicated security-engineer subagent that inspects source files across six audit categories, classifies findings by severity, and generates a structured markdown report for prioritized remediation.

To integrate a security remediation report into a release gate, generate a structured markdown report classifying findings by severity, which becomes a required artefact for the Polish to Release gate before public release or multiplayer launch.