Simple Weather MCP Server

Overview

The Mcp Client Server implementation provides a lightweight, end‑to‑end example of how to expose and consume data through the Model Context Protocol (MCP). It solves a common developer pain point: integrating external, structured data sources into AI assistants without writing custom adapters for each assistant. By packaging weather information as MCP resources, tools, and prompts, the server turns raw data into a first‑class citizen that can be queried, manipulated, and combined with other AI content in a single conversation.

What the Server Does

At its core, the server hosts three types of MCP entities:

• Resources that expose static city data (, , ).
• Tools that perform dynamic operations, such as fetching a weather forecast for any city over an arbitrary number of days.
• Prompts that generate contextual advice, e.g., travel recommendations based on the forecast.

These entities are accessible through JSON‑RPC 2.0 over standard I/O streams, allowing any MCP‑compatible client—whether a local script or an AI assistant like Claude—to list available resources, read specific data, invoke tools with parameters, and retrieve prompt results. The server is intentionally simple: it focuses on demonstrating the protocol rather than providing a production‑ready weather API, yet its design is fully extensible.

Why It Matters for AI Developers

Integrating external data into conversational agents often requires bespoke APIs, authentication handling, and custom parsing logic. MCP abstracts these concerns by presenting a uniform interface: resources for static data, tools for actions, and prompts for context generation. Developers can therefore:

• Rapidly prototype new data sources by adding a few JSON objects or function stubs.
• Maintain consistency across different assistants, as the MCP contract is language‑agnostic.
• Debug efficiently, thanks to the JSON‑RPC logs that expose every request and response in a human‑readable format.

The example server also demonstrates how to spawn the MCP process as a child, stream data over stdio, and log communication—all patterns that are common in real‑world deployments.

Key Features Explained

• Resource Listing & Retrieval – Clients can enumerate all available city resources and fetch detailed information in a single call, enabling dynamic UI generation or data caching.
• Parameterized Tool Calls – The tool accepts a city name and day count, illustrating how to expose complex operations while keeping the interface simple.
• Prompt Generation – The prompt shows how to transform raw data into natural language recommendations, bridging the gap between structured input and conversational output.
• JSON‑RPC Transparency – By exposing raw JSON‑RPC messages in the console, developers gain insight into request/response flows and can trace issues quickly.
• Transport Flexibility – Although this demo uses stdio, the underlying MCP SDK supports other transports (e.g., TCP, WebSocket), making it easy to migrate to production environments.

Real‑World Use Cases

• Travel Planning Bots – A travel assistant can query the tool for a destination and then use to suggest packing lists or activities.
• Smart Home Integrations – A home automation system can expose room data as resources and trigger climate control tools based on weather forecasts.
• Data‑Driven Analytics – Analysts can pull city resources and forecast data into a single pipeline, enabling automated reporting or anomaly detection.
• Educational Tools – Students can experiment with MCP by adding new resources (e.g., historical events) and building custom prompts that weave facts into storytelling.

Standout Advantages

• Zero Boilerplate – The example requires no external API keys or complex setup; the entire weather dataset is bundled locally.
• Clear Separation of Concerns – Resources, tools, and prompts are distinct, making it straightforward to add new capabilities without touching existing logic.
• Extensible Transport – The stdio implementation is a lightweight starting point, but the same code can be adapted to network sockets or cloud functions with minimal changes.
• Developer Friendly Logging – The custom logging transport gives instant visibility into the protocol dance, reducing friction during debugging.

In summary, the Mcp Client Server demonstrates how MCP can turn any data source into a first‑class conversational asset, streamlining the integration of structured information into AI assistants and enabling developers to focus on value‑adding logic rather than plumbing.