ce-agent-native-audit

To audit your codebase for agent-native architecture alignment, apply this Skill to repositories containing agent-native tooling to generate per-principle scores across eight defined areas and a consolidated markdown report with prioritized recommendations.

An agent-native architecture review is a structured assessment that scores a codebase against eight defined principles to identify alignment gaps. It uses parallel sub-agents to discover architectural metrics and surfaces actionable recommendations for governance and improvement.

This architecture audit requires repositories containing agent-native tooling or components. Applying it to standard codebases without agent-native components will not yield meaningful per-principle scores because the eight discovery sub-agents specifically target agent-native architecture patterns.

To generate an architecture scorecard, invoke the audit on your repository to trigger eight parallel discovery sub-agents. These sub-agents evaluate the codebase against defined principles, producing per-principle scores and a consolidated markdown summary with risk indicators and prioritized recommendations.

The best way to review agent-native architecture for governance is running a full audit that produces a consolidated report. This multi-agent workflow evaluates eight architecture principles, generating actionable recommendations and risk indicators to guide codebase improvements.

When running a multi-agent architecture scorecard on a large codebase, the audit requires deploying eight parallel discovery sub-agents, which may consume significant processing resources. The final markdown report focuses strictly on eight defined principles, potentially missing unrelated architectural risks outside that scope.