mcp-hko
Fetch data from Hong Kong Observatory with MCP
Documentation
mcp-hko (Hong Kong Observatory MCP Server)
This MCP server provides access to Hong Kong Observatory weather data APIs.
It implements various tools that allow clients to retrieve weather forecasts, earthquake information, tide data, and other meteorological information through the Model Context Protocol.
Available Weather Data Tools
- 9-day Weather Forecast (fnd)
- Current Weather Report (rhrread)
- Local Weather Forecast (flw)
- Weather Warning Information (warningInfo)
- Weather Warning Summary (warnsum)
- Special Weather Tips (swt)
- Quick Earthquake Messages (qem)
- Locally Felt Earth Tremor Report (feltearthquake)
- Tide Information (HHOT, HLT)
- Astronomical Data (SRS, lunardate)
- Lightning and Visibility Data (LHL, LTMV)
- Temperature Records (CLMTEMP, CLMMAXT, CLMMINT)
- Radiation Level Reports (RYES)
- Rainfall Data (hourlyrainfall)
Usage Examples
Get 9-day weather forecast
> prompt:
>
> what is the weather in next Wed ?
Get current weather report
> prompt:
>
> what is the weather now ?
Development Setup
1. Install dependencies:
npm install2. Start the MCP development server:
npm run inspect3. Build the MCP server:
npm run build4. Run tests:
npm test5. Run tests in watch mode:
npm run test:watchComprehensive tests have been created for all API functions. See docs/TESTING.md for details.
Configuration (Claude Desktop)
{
"mcpServers": {
"mcp-hko": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"github:louiscklaw/mcp-hko"
]
}
}
}Configuration (docker)
{
"mcpServers": {
"mcp-hko-docker": {
"command": "docker",
"args": ["run", "-i", "--rm", "mcp-hko:latest"],
"name": "mcp-hko (Docker)",
}
}
}Project Structure
- `src/` - Server source code
- `src/lib/` - API implementation modules
- `documentation/` - API documentation and test cases
- `scripts/` - Development utilities
- `src/tests/` - Test files
Server Architecture
The MCP server is built on the FastMCP framework and implements various tools that correspond to Hong Kong Observatory APIs.
The architecture is shown in two diagrams for better readability:
Server Overview
Detailed Tool Structure
REQ01xx: Weather Forecasts & Warnings
flowchart TD
subgraph REQ01xx[Weather Forecasts & Warnings]
direction BT
B1[9-day Forecast]
B2[Current Weather]
B3[Local Forecast]
B4[Warning Info]
B5[Warning Summary]
B6[Special Weather Tips]
endREQ02xx: Earthquake Information
flowchart TD
subgraph REQ02xx[Earthquake Information]
C1[Quick Earthquake]
C2[Felt Earthquake]
endREQ03xx: Tides, Astronomy & Climate
flowchart TD
subgraph REQ03xx[Tides, Astronomy & Climate]
D1[Tide: HHOT]
D2[Tide: HLT]
D3[Sunrise/Sunset: SRS]
D4[Lightning: LHL]
D5[Visibility: LTMV]
D6[Temperature: CLMTEMP]
D7[Max Temperature: CLMMAXT]
D8[Min Temperature: CLMMINT]
D9[Radiation: RYES]
endREQ04xx: Lunar Calendar
flowchart TD
subgraph REQ04xx[Lunar Calendar]
E1[Lunar Date]
endREQ05xx: Rainfall
flowchart TD
subgraph REQ05xx[Rainfall]
F1[Hourly Rainfall]
endCredits:
Frequently asked questions
What is mcp-hko?
mcp-hko is Fetch data from Hong Kong Observatory with MCP
How do I install mcp-hko?
Open the GitHub repository and follow its README. Most MCP servers are added to your client's MCP config, then called by your agent.
Is mcp-hko open source?
Yes — it is hosted on GitHub at https://github.com/louiscklaw/hko-mcp and has 1 stars.
Related MCP tools
Directory for Awesome MCP Servers TypeScript-based implementation. Trusted by 1900+ developers. Trusted by 1900+ developers.
MCP Aggregator, Orchestrator, Middleware, Gateway in one docker TypeScript-based implementation. Trusted by 1400+ developers.
A Unified MCP Server Management App TypeScript-based implementation. Trusted by 800+ developers. Trusted by 800+ developers.
MCP Server for SearXNG TypeScript-based implementation.
Enhanced ChatGPT Clone: Features Agents, MCP, DeepSeek, Anthropic, AWS, OpenAI, Responses API, Azure, Groq, o1, GPT-5, Mistral, OpenRouter, Vertex AI, Gemini...
FastGPT is a knowledge-based platform built on the LLMs, offers a comprehensive suite of out-of-the-box capabilities such as data processing, RAG retrieval, ...
Run your own MCP server? See who uses it and what to fix.
Measure it with TrackMCP