Skip to content

SE Theory: Reference Kit

se-theory-reference-kit is shared Python infrastructure for Structural Explainability theory repositories.

It provides generic tooling for maintaining theory-reference artifacts that mirror Lean public surfaces. The kit supports loading, inspecting, validating, scaffolding, exporting, and cataloging reference material supplied by the owning theory repository.

Purpose

The package exists so Structural Explainability theory repositories can share one reference-tooling engine instead of each maintaining separate Python tooling for the same repository mechanics.

The kit serves theory repositories that maintain Lean source and repo-owned reference artifacts. It does not define theory content.

Scope

The kit owns generic infrastructure for:

  • repository-relative path handling,
  • typed declaration loading,
  • Lean source inspection,
  • reference artifact loading,
  • reference registry construction,
  • reference validation,
  • scaffold support,
  • generated export support,
  • catalog support,
  • command orchestration.

The kit does not own:

  • Lean semantics,
  • formal definitions,
  • public symbol lists,
  • theory-specific export maps,
  • contract packaging,
  • operational validation pipelines,
  • domain mappings,
  • runtime systems.

Repository ownership boundary

Each theory repository owns its own declared public Lean surface and export specifications.

The shared kit consumes those declarations through typed repository-provided objects. This keeps theory content in the theory repository and keeps the kit focused on reusable mechanics.

Public command

The package exposes one public command:

se-theory-reference

See Commands for the command responsibilities.

API reference

The Python API reference is generated from source during documentation builds.

See API Reference for generated package documentation.