Skip to content
AAOE
Draft3 min read

Context Engineering

Observation

During the SolOh ERP Modernization and the evolution of AAOE itself, better organized engineering context consistently produced higher-quality AI implementation. AI agents required progressively fewer prompt instructions as repository knowledge matured. Shared terminology reduced ambiguity across specialized AI agents. Stable specifications reduced prompt complexity. Repository organization influenced engineering quality.

Context

These observations occurred throughout the SolOh ERP Modernization and the resulting evolution of the AAOE repository itself, across specialized AI agents, Work Orders, the growing Discovery Catalog, and specification-first engineering more broadly. CS-0001 — SolOh ERP Modernization is the primary evidence source for this Discovery.

Evidence

  • Prompts became shorter while implementation quality improved.
  • AI agents increasingly relied on repository artifacts instead of prompt instructions.
  • Shared specifications reduced repeated explanations across Work Orders.
  • Architectural consistency improved as documentation expanded.
  • Repository structure became part of the engineering process itself.

Discovery

Engineering context functions as reusable infrastructure for AI-assisted software engineering. The quality of that context correlates more strongly with engineering outcomes than prompt sophistication alone.

Why It Matters

This observation appears generally applicable beyond SolOh:

  • Scalability — context accumulated once can be drawn on by many subsequent Work Orders, rather than being rebuilt each time.
  • Repeatability — implementation quality became less dependent on how any individual prompt was phrased.
  • Specialization — specialized agents could rely on shared context instead of needing identical prompt histories.
  • Organizational knowledge — context, once recorded, persists beyond the task that produced it.
  • Onboarding — new contributors, human or agent, could draw on the same accumulated context rather than starting from nothing.
  • Consistency — shared terminology and specifications reduced divergence between independently executed Work Orders.
  • Traceability — context artifacts such as specifications, ADRs, and Discovery records connect implementation back to the reasoning behind it.

Consequences

  • Engineering knowledge becomes a reusable asset.
  • Prompts become lightweight execution mechanisms.
  • Repository organization becomes part of engineering infrastructure.
  • AI agents collaborate through shared context instead of shared prompt history.
  • Organizations accumulate engineering capability through documented context.

Case Studies

This Discovery is referenced by, and later generalized into, the following AAOE specifications:

It may also motivate a future Context Engineering Specification, not yet published in this repository.

Future Evolution

Future projects may validate, refine, or challenge this Discovery. Engineering Discoveries evolve through accumulated evidence, not through revision of this record's original observation.