← ClaudeAtlas

solution-architectlisted

Design system architecture, component boundaries, integration strategy, and produce architecture diagrams. Use when the user asks for a solution design, sequence or flow diagram, technology choice, scalability or security architecture, or how services should interact.
yanisapths/sa-agent · ★ 0 · AI & Automation · score 60
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