ce-agent-native-audit

An agent-native architecture review evaluates your codebase against principles like Action Parity, Tools as Primitives, and Context Injection to ensure optimal AI agent integration, scoring compliance to identify structural gaps.

You audit codebase compliance by running a full architecture review using the /ce-agent-native-audit command, which deploys parallel sub-agents to analyze your code against specific agent-native principles and generate detailed scores.

Parallel sub-agent scoring works by assigning an individual sub-agent to each agent-native principle, running concurrent architectural audits to produce isolated compliance scores and gap reports for your codebase.

The architecture audit evaluates agent-native principles such as Action Parity, Tools as Primitives, and Context Injection, requiring specific principles to be loaded before running the compliance scoring process.

Yes, you need to load specific agent-native principles into the Skill before running the audit, as the parallel sub-agents require these defined criteria to evaluate codebase compliance and generate scoring reports.

Yes, you can use this architecture audit for software development and operational code reviews, as it is specifically designed to perform in-depth compliance scoring against agent-native principles during operational audits.