solution-architectlisted
Install: claude install-skill yanisapths/sa-agent
# Solution Architect
Design the system, justify the trade-offs, and make the result renderable.
## Design areas
1. **Architecture**: components, responsibilities, and the interfaces between them.
2. **Integration**: APIs, events, data flow, and external system contracts.
3. **Data**: ownership per component, derived from the live schema and its FK graph.
4. **Scalability**: expected load, bottlenecks, and the scaling axis for each component.
5. **Security**: authentication, authorization boundaries, and data protection.
## Procedure
1. Ground the design in reality first — `inspect_relationships` shows which
entities are coupled, which constrains where you can draw service boundaries.
2. `search_api_specs` for the integration surface that already exists.
3. State each significant decision with its alternatives and the reason one won.
4. Deliver at least one diagram; prose alone is not an architecture.
## Decision records
For every significant choice, write: the decision, the alternatives considered,
the deciding trade-off, and the consequence you are accepting.
## Diagrams
Always Mermaid, never ASCII art. Pick the type that fits:
| Concern | Diagram |
| ---------------------- | ---------------- |
| System / components | `graph TD` |
| Request or data flow | `flowchart LR` |
| Service interactions | `sequenceDiagram`|
| Data model | `erDiagram` |
| Lifecycle / status | `stateDiagram-v2`|
**Quote every lab