Skip to main content
Glama
mlaurel

Structured Workflow Engine MCP Server

by mlaurel
architecture-analysis.md4.79 kB
# Architecture Analysis Prompt ## 🎯 Goal Describe the current architecture, highlight weak spots, propose improvements, and add analysis to the TRD document—no fluff. ## 📥 Context (ask if missing) 1. **Scope** – which services/modules to inspect. 2. **Codebase** – repo paths or file access. 3. **Docs** – diagrams, ADRs, design notes. 4. **Focus** – pick the priority: scalability / performance / security / maintainability. 5. **Feature Name** – for TRD filename (e.g., "user-authentication", "payment-processing") 6. **Existing TRD** – check if `docs/planning/[feature-name]-trd.md` exists ## 🚦 Skip the analysis if - The system is trivial **and** unchanged. - A fresh report exists (< 30 days old). - Only an urgent bug-fix is needed, with no architecture impact. ## 🔍 Analysis Checklist - [ ] Components + responsibilities - [ ] Data flows & dependencies - [ ] Architectural patterns - [ ] Data stores, consistency, access patterns - [ ] Quality attributes: ⬆️ scalability, ⚡ performance, 🔒 security, 🛠️ maintainability ## 📤 Output 1. Run architecture analysis on the codebase 2. **Add/Update in TRD:** `docs/planning/[feature-name]-trd.md` ### TRD Section Structure Add the following section to the TRD document: ```markdown ## 5. Architecture Analysis ### 5.1 System Overview **Architecture Style:** [Monolith/Microservices/Serverless/Hybrid] **Primary Patterns:** [MVC, Event-driven, CQRS, etc.] ### 5.2 Component Architecture ```mermaid graph TB A[Component A] --> B[Component B] B --> C[Component C] A --> D[External Service] ``` **Component Responsibilities:** - **Component A:** [Responsibility and files] - **Component B:** [Responsibility and files] - **Component C:** [Responsibility and files] ### 5.3 Data Architecture **Data Stores:** - **Primary Database:** [Type, purpose, access patterns] - **Cache Layer:** [Type, purpose, data cached] - **External Data:** [APIs, third-party sources] **Data Flow:** ```mermaid flowchart LR A[User Input] --> B[API Layer] B --> C[Business Logic] C --> D[Data Layer] D --> E[Database] ``` **Consistency Model:** [ACID/Eventual consistency/etc.] ### 5.4 Integration Architecture **External Dependencies:** - [Service 1: Purpose, SLA, failure handling] - [Service 2: Purpose, SLA, failure handling] **API Design:** - **Style:** [REST/GraphQL/gRPC] - **Authentication:** [Method used] - **Rate Limiting:** [Current implementation] ### 5.5 Quality Attributes Assessment #### 🔒 Security **Strengths:** - [Security strength 1] - [Security strength 2] **Weaknesses:** - [Security gap 1: Impact and recommendation] - [Security gap 2: Impact and recommendation] #### ⚡ Performance **Current Performance:** - **Response Times:** [Measurements] - **Throughput:** [Requests/sec capacity] - **Bottlenecks:** [Identified performance issues] **Optimization Opportunities:** - [Performance improvement 1: Impact estimate] - [Performance improvement 2: Impact estimate] #### ⬆️ Scalability **Current Capacity:** - **Horizontal Scaling:** [Current capabilities] - **Vertical Scaling:** [Current limits] - **Auto-scaling:** [Implementation status] **Scalability Constraints:** - [Constraint 1: Description and impact] - [Constraint 2: Description and impact] #### 🛠️ Maintainability **Code Organization:** - **Modularity:** [Assessment of separation of concerns] - **Coupling:** [Tight/loose coupling assessment] - **Cohesion:** [Component cohesion evaluation] **Technical Debt:** - [Debt item 1: Impact on maintenance] - [Debt item 2: Impact on maintenance] ### 5.6 Architecture Improvements **High Priority:** - [Improvement 1: Description, impact, effort] - [Improvement 2: Description, impact, effort] **Medium Priority:** - [Improvement 3: Description, impact, effort] **Future Considerations:** - [Long-term architectural evolution] - [Technology upgrade path] ### 5.7 Architecture Decision Records (ADRs) **Key Decisions:** - **Decision 1:** [What was decided and why] - **Decision 2:** [What was decided and why] **Trade-offs Made:** - [Trade-off 1: What was sacrificed for what benefit] - [Trade-off 2: What was sacrificed for what benefit] ``` **Note:** Include architecture diagrams using Mermaid. Reference specific components, services, and integration points with file paths where applicable. ## ➡️ Response Flow ```mermaid flowchart LR A[User] -->|request| B[Analysis Engine] B --> C{TRD exists?} C -- Yes --> D[Read existing TRD] C -- No --> E[Create new TRD] D --> F{Need more context?} E --> F F -- Yes --> G[Ask clarifying questions] F -- No --> H[Run architecture analysis] G --> H H --> I[Add Architecture Analysis section] I --> J[Update TRD file] ```

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/mlaurel/mcp-workflow-engine'

If you have feedback or need assistance with the MCP directory API, please join our Discord server