Summary

1. Model for an engineering question. A model is a purposeful reduction of a system. Preserve the distinctions needed to answer the question; omit detail that does not contribute to it. Build and maintain the model only while the engineering leverage it provides justifies its carrying cost.
2. Externalize engineering knowledge and intent. Models can expose architecture, obligations, dependencies, invariants, provenance, costs, and other properties that would otherwise have to be reconstructed from implementation and context. Different questions may require different models of the same system.
3. Choose degrees of freedom deliberately. Every representation leaves some choices open. Reduce consequential freedom where doing so improves reasoning or control; preserve freedom where realization can safely remain open.
4. Maintain the correspondences you depend on. Multiple representations create correspondence obligations. Derivation, generation, validation, shared identity, and other mechanisms can keep related representations sufficiently consistent for their intended uses.
5. Distinguish correspondence from governance. A model may describe another artifact, or realization may be required to follow the model. Decide which representation owns each consequential fact and which consumers must follow it.

Modeling principle

Externalize engineering knowledge and intent into explicit, structured models that both engineers and agents can reason through.

Appropriate representations make broader engineering questions tractable and make additional properties available for Alignment.