mcp-architecture-flow.md•4 kB
# MCP ADR Analysis Server Architecture Flow
This document shows the system architecture and data flow using Mermaid diagrams.
## System Architecture Overview
```mermaid
graph TD
A[AI Assistant<br/>Claude/Cursor/Cline] --> B[MCP Client]
B --> C[MCP ADR Analysis Server]
C --> D[Project Analysis Engine]
C --> E[AI Execution Engine]
C --> F[Cache Management]
D --> G[File System Scanner]
D --> H[Technology Detector]
D --> I[Pattern Analyzer]
E --> J[OpenRouter API]
J --> K[Claude-3-Sonnet]
J --> L[GPT-4o]
C --> M[ADR Generator]
C --> N[Security Scanner]
C --> O[Deployment Validator]
M --> P[././adrs/]
N --> Q[Content Masking]
O --> R[Health Scoring]
style A fill:#e1f5fe
style C fill:#fff3e0
style J fill:#f3e5f5
style P fill:#e8f5e8
```
## Tool Execution Flow
```mermaid
sequenceDiagram
participant User as AI Assistant User
participant AI as AI Assistant
participant MCP as MCP Client
participant Server as ADR Analysis Server
participant Engine as AI Execution Engine
participant API as OpenRouter API
User->>AI: "Analyze this React project"
AI->>MCP: Tool call: analyze_project_ecosystem
MCP->>Server: MCP Protocol Request
alt AI Execution Mode
Server->>Engine: Process with AI enhancement
Engine->>API: Send enhanced prompts
API-->>Engine: AI analysis results
Engine-->>Server: Processed insights
else Prompt-Only Mode
Server->>Server: Generate analysis prompts
end
Server-->>MCP: Analysis results/prompts
MCP-->>AI: Tool response
AI-->>User: Actionable insights
```
## Decision Flow Architecture
```mermaid
flowchart LR
A[Project Input] --> B{Existing ADRs?}
B -->|Yes| C[discover_existing_adrs]
B -->|No| D[suggest_adrs]
C --> E[compare_adr_progress]
D --> F[generate_adr_from_decision]
E --> G{Implementation Gap?}
F --> G
G -->|Yes| H[generate_adr_todo]
G -->|No| I[deployment_readiness]
H --> J[manage_todo_json]
I --> K{Ready for Deploy?}
K -->|Yes| L[smart_git_push]
K -->|No| M[Address Issues]
M --> N[get_workflow_guidance]
N --> H
style A fill:#e3f2fd
style L fill:#e8f5e8
style M fill:#fff3e0
```
## Security Analysis Flow
```mermaid
graph TB
A[Source Code] --> B[analyze_content_security]
B --> C{Sensitive Content Found?}
C -->|Yes| D[generate_content_masking]
C -->|No| E[Security Clear ✅]
D --> F[configure_custom_patterns]
F --> G[apply_basic_content_masking]
G --> H[validate_content_masking]
H --> I{Validation Passed?}
I -->|Yes| J[Content Secured ✅]
I -->|No| K[Update Patterns]
K --> F
style A fill:#f3e5f5
style E fill:#e8f5e8
style J fill:#e8f5e8
style K fill:#ffebee
```
## Deployment Readiness Pipeline
```mermaid
stateDiagram-v2
[*] --> Analyzing
Analyzing --> SecurityScan: analyze_content_security
SecurityScan --> SecurityPassed: No issues found
SecurityScan --> SecurityFailed: Issues detected
SecurityFailed --> Masking: generate_content_masking
Masking --> SecurityScan: Re-validate
SecurityPassed --> TestValidation: deployment_readiness
TestValidation --> TestsPassed: All tests pass
TestValidation --> TestsFailed: Tests failing
TestsFailed --> FixRequired: Address issues
FixRequired --> TestValidation: Re-run validation
TestsPassed --> DeploymentReady: smart_git_push
DeploymentReady --> [*]: # Deployment Complete
SecurityFailed --> [*]: # Manual Intervention Required
FixRequired --> [*]: # Development Needed
```
These diagrams illustrate the comprehensive workflow and architecture of the MCP ADR Analysis Server, showing how different components interact to provide AI-powered architectural analysis and decision support.