You are an expert AI systems architect specializing in Model Context Protocol (MCP) server development (https://modelcontextprotocol.io/). Create a comprehensive, detailed implementation plan for the following MCP server:
[Provide a concise description of your MCP server concept - what it does and the primary value it delivers]
- Example: "This MCP server enables real-time context sharing between AI models and client applications, improving response accuracy and user interaction efficiency."
- Who will interact with this MCP server?
[Types of users, e.g., developers building AI applications, AI agents, end-users via apps] - What are their primary goals?
[User objectives, e.g., seamless model context integration, reduced latency in AI responses] - What problems does this solve for them?
[Issues addressed, e.g., lack of standardized context management, inefficient data exchange]
Provide complete details for ALL required capabilities of the MCP server:
- Capability 1: [Description with complete requirements, e.g., "Context ingestion from client inputs with support for text and metadata"]
- Capability 2: [Description with complete requirements, e.g., "Real-time context synchronization across connected models"]
- Capability 3: [Description with complete requirements, e.g., "Context retrieval API for external applications"]
- [Continue with all needed capabilities]
- Interaction style:
[e.g., REST API, WebSocket, GraphQL - specify how clients connect to the MCP server] - Protocol requirements:
[Specific protocols or standards, e.g., JSON-RPC over HTTPS, WebSocket with custom framing] - Key components needed:
[e.g., API endpoints like/context/create
, message formats like{ "context_id": "xyz", "data": {...} }
]
- Describe all entities and their relationships:
[Entities and connections, e.g., "Context (ID, data, timestamp) linked to Session (user_id, session_id)"] - Specify data types and validation rules:
[Data specifications, e.g., "Context data: JSON object, max size 1MB; Timestamp: ISO 8601 string"] - Include authentication/authorization model if applicable:
[Security model, e.g., "OAuth 2.0 tokens required for all requests"]
Detail the complete user journey when interacting with the MCP server:
- Entry point: [How users or agents connect, e.g., "Client sends POST to
/context/init
"] - Authentication: [If applicable, e.g., "Server validates JWT in Authorization header"]
- Main interaction: [Step-by-step flow, e.g., "Client submits context data, server processes and assigns context_id"]
- Edge cases: [How to handle failures/exceptions, e.g., "Invalid data returns 400 Bad Request"]
- Success states: [What completion looks like, e.g., "Server returns 200 OK with context_id"]
- MCP Framework:
[Specify the MCP framework to use, e.g., Agno, JADE, SPADE, MESA, or a custom implementation based on modelcontextprotocol.io] - Backend requirements:
[If applicable, e.g., "Node.js with Express, hosted on AWS EC2"] - APIs/Integrations needed:
[List all external services, e.g., "Integration with Redis for caching"] - Performance expectations:
[Throughput/Response times, e.g., "Handle 1000 requests/sec with <50ms latency"] - Security considerations:
[User authentication/Data protection, e.g., "TLS 1.3 encryption, rate limiting"]
- Required accessibility standards:
[Interaction standards, if applicable, e.g., "API documentation compliant with OpenAPI 3.0"] - Compliance requirements:
[GDPR/CCPA/etc if applicable, e.g., "User data anonymization per GDPR"]
- Similar MCP servers:
[Descriptions or references, e.g., "Context management in Hugging Face’s inference server"] - Specific elements to emulate:
[Capabilities/Interactions, e.g., "Low-latency context updates like Google’s Realtime API"]
- Must-have capabilities:
[Critical functionality, e.g., "Context ingestion and retrieval"] - Nice-to-have capabilities:
[Secondary importance, e.g., "Support for multi-language contexts"] - Future enhancements:
[Post-initial release, e.g., "Context versioning system"]
[Any other information that would help create the best possible implementation, e.g., "Target deployment by Q1 2025 with a team of 3 developers"]