Bilingual navigation: Español
This area contains the reusable architecture model. Read it from general policy to concrete evidence.
Goal: define the reusable, provider-neutral architecture model that every product inherits — from baseline principles down to runtime patterns.
Objectives:
- Keep the runtime-agnostic baseline and blueprints as the single source of architectural policy.
- Record every accepted trade-off as an ADR, discoverable through the decision matrix.
- Provide runtime-conditioned implementation guidance without turning it into universal policy.
Ordered from general policy to concrete evidence:
| Document | Description | Goal / Objective | Type | Mandatory |
|---|---|---|---|---|
| Principles | Foundational architecture principles | Ground all decisions in shared principles | Area hub | Yes |
| Blueprints Hub | Runtime-agnostic principles, topology, and selection criteria | Define the architectural baseline | Area hub | Yes |
| ADR Registry | Records accepted trade-offs and their scope | Preserve decision history | Area hub | Yes |
| ADR Matrix | Finds controlling ADRs by architectural concern | Speed up decision discovery | Decision index | Yes |
| Canonical Patterns | Code patterns governed by runtime-specific ADRs | Standardize runtime implementations | Area hub | No |
| Evolith SDK | Domain model and technical design of the Evolith SDK | Design the shared SDK | Design reference | No |
| MCP Tools Catalog | Catalog of 11 MCP tools for AI agent automation | Enable AI agent automation | Tool reference | No |
| UMS Reference Model | Shows how a real product adopts or specializes the reference | Demonstrate applied evidence | Applied reference | No |
The baseline and accepted ADRs define policy. Runtime profiles and canonical patterns define conditioned implementation guidance. UMS is the official applied reference model: it demonstrates decisions in a complete product, but its product-specific choices are not automatically universal standards.