golden-principles

AI agent operating principles are rules that govern agent behavior, context management, and adherence to best practices like using the repository as the system of record to ensure secure infrastructure workflows.

To enforce mechanical enforcement over documentation for AI agents, apply core operating principles that establish invariants and govern how agents interact with the repository, ensuring rules are executed systematically rather than just documented.

AI agents manage context by consulting operating principles to store new information in designated repository files rather than relying on ephemeral chat history, ensuring the repository remains the system of record.

You need to define agent governance principles for Azure infrastructure workflows when you want to ensure consistent, secure, and efficient development by establishing rules for how agents interact with repository data and enforce invariants.

These agent operating principles are designed for the Azure Agentic InfraOps Accelerator repository, but the core governance patterns like repository as system of record can be applied to infrastructure development workflows in other repositories.

Using the repository as the system of record instead of chat history ensures that AI agents store new information in designated files, maintaining persistent context management and consistent infrastructure development workflows.