my-mcp-server
# my-mcp-server
A simple Model Context Protocol (MCP) server implementation in Node.js. This server exposes three tools via MCP:
- **greet**: Greet someone with a personalized message.
- **calculate**: Perform basic math calculations.
- **weather_info**: Get mock weather information for a city.
Above ones are implemented in the server1.js
in server2 there is an agent that requires selenium web driver that launches the html provided in prompt in a browser and tests and returns if any console errors are thrown.
All these agents can be used in not just copilot but with other AI extensions too for example refer this video for Amazon Q --> https://www.youtube.com/watch?v=sF6KkaV102M
## Features
- MCP server using `@modelcontextprotocol/sdk`
- Stdio transport for easy integration
- Extensible tool definitions
## Steps
1. **Clone the repo and install dependencies**
```powershell
git clone https://github.com/devashish234073/my-mcps
cd my-mcps
npm install
```
2. Open the my-mcp folder in vscode
3. Open copilot chat
<br><img width="508" height="308" alt="image" src="https://github.com/user-attachments/assets/5cdec5fd-5931-48b4-8e28-e94ba1342e9c" />
4. In the chat click on "Configure Tools"
<br><img width="653" height="343" alt="image" src="https://github.com/user-attachments/assets/b819df14-72c7-407b-911f-43169b7d57b3" />
5. Add more tools
<br><img width="647" height="329" alt="image" src="https://github.com/user-attachments/assets/1a1487d6-3bb7-483a-956f-ea10b82dc7cd" />
6. Add MCP Server
<br><img width="669" height="168" alt="image" src="https://github.com/user-attachments/assets/f80488df-9a40-4cfb-a25d-c2ea8791d53d" />
7. This creates an mcp.json file with minimal configuration of the command your mcp server will run, in our case the repo already has that so step #4 to #7 is not needed
<br><img width="572" height="112" alt="image" src="https://github.com/user-attachments/assets/f80488df-9a40-4cfb-a25d-c2ea8791d53d" />
<br><img width="572" height="227" alt="image" src="https://github.com/user-attachments/assets/0f2e8946-5101-48dc-9f7e-160b89597383" />
8. Just go to that file and click "Restart"
9. The server.js has three tools greet, calculate and weather_info
<br><img width="432" height="242" alt="image" src="https://github.com/user-attachments/assets/db22b737-cde4-4950-8885-e24b965c2c2a" />
10. You can interact with these now using prompts like:
<br><img width="330" height="550" alt="image" src="https://github.com/user-attachments/assets/f9e24139-89e7-4630-8bd5-d61b80f67fb2" />
TDQS
Scored across 3 tools
Each tool has a clearly distinct and non-overlapping purpose: calculate handles math, greet handles messaging, and weather_info handles weather data. An agent can easily tell them apart with no risk of misselection.
The naming is mixed: calculate and greet are verbs, while weather_info is a noun phrase. There is no consistent verb_noun pattern, but the names are still readable and descriptive.
With only 3 tools, the server feels thin and under-scoped for a general-purpose MCP server. The tools cover unrelated domains (math, greeting, weather), suggesting a lack of cohesive focus.
The server lacks a clear domain, making completeness hard to assess. Each tool stands alone without supporting operations (e.g., no update or delete for weather), leading to significant gaps in any potential workflow.