gemini-code-auditor

To audit production code for security weaknesses, apply structured adversarial review prompts against root entrypoints and critical subsystems. This process surfaces line-level issues, race conditions, and misconfigurations by inspecting initialization cascades and startup ordering.

An adversarial code review is a technique that uses structured prompts to challenge code correctness and systematically expose security gaps, permission issues, and hard-to-detect logic flaws with precise line references. It targets non-obvious architectural weaknesses.

Generate line-level remediation guidance by running an adversarial audit that outputs severity-ranked findings with exact code references and 3-line pseudo-fixes. This quantified risk reporting provides concrete hints to guide immediate remediation of structural flaws.

Yes, you can use adversarial review prompts to detect TOCTOU checks and silent exception swallowing. These are known risk patterns specifically targeted during the systematic analysis of critical subsystems and initialization cascades.

The best way to audit startup ordering and initialization cascades is to apply structured adversarial prompts directly to root entrypoints. This systematically inspects primary orchestrators to uncover init-time failures and architectural misconfigurations.