research-document

Repository Architecture

Repository Architecture

Generated: 2026-07-22T15:10:59+00:00

Recommendation

Use a hybrid architecture:

  • Concept-first for knowledge relationships
  • Project-first for active research programs
  • Artifact-type collections for generated indexes and registries
  • Generated data layers for search, graph traversal, and website navigation

Why This Fits The Evidence

The repository has 49 files but already spans multiple projects, document types, and maturity levels. A discipline-first tree would fragment shared concepts. A project-first tree would hide cross-project laws and evidence. An artifact-first tree would optimize storage at the expense of reasoning. A hybrid model keeps canonical concepts stable while allowing projects and derived artifacts to evolve independently.

Target Structure

content/
  concepts/
    perception/
    attention/
    hierarchy/
    wayfinding/
    color/
    typography/
  projects/
    composition-science/
      governance/
      canonical/
      research/
      evidence/
    project-atlas/
      charter/
      canonical/
      reports/
      evidence/
    product-genome/
      canonical/
      reports/
  registries/
    evidence/
    hypotheses/
    experiments/
    decisions/
  journals/
    research-journal/
data/
  catalog/
  graph/
  search/
  lineage/
site/
scripts/

Canonical Rules

  • Canonical concept pages own definitions and stable identifiers.
  • Derived documents never duplicate canonical definitions; they cite them.
  • Registries are generated, not edited by hand.
  • Historical documents remain preserved under lineage-aware versioning.