mcp-reminder
If you are the rightful owner of mcp-reminder and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.
Reminder MCP Server with Spring+OpenAPI+SQLite for Open WebUI is designed to store reminders in a SQLite database, utilizing OpenAPI for REST API endpoint descriptions.
Reminder MCP Server with Spring+OpenAPI+SQLite for Open WebUI
saveReminder + getCurrentTime
In this application, we store reminders in a SQLite database. Before saving, we always call getCurrentTime to retrieve the current date and time. This ensures that any relative dueDate (like “tomorrow” or “next Friday”) is calculated correctly in relation to the actual current date and time.
Explanation
Open WebUI implements MCP through the OpenAPI definitions of our REST API.
We build a common RestAPI but use Swagger+OpenAPI to describe each endpoint exposes as a Tool.
For example:
@Operation(
summary = "Sum two numbers",
description = """
This tool takes two numbers as input (`a` and `b`) and returns their sum.
When this tool is available, the model can perform addition by calling this function instead of calculating internally.
Input:
- `a`: first integer
- `b`: second integer
Output:
- `result`: the sum of `a + b`
- `ok`: a boolean indicating success
Example:
Request: { "a": 3, "b": 5 }
Response: { "ok": true, "result": 8 }
"""
)
@PostMapping
public ResponseEntity<Object> saveReminder(int a,int b)
{
return ResponseEntity.internalServerError()
.body(Map.of(
"ok",true,
"result",a+b
));
}
Open WebUI integration
Now that our REST API (the "MCP") is ready, here's how to connect it in Open WebUI.
- In Open WebUI, go to Settings > Tools > Manage Tool Servers.
- Click Add Server (or Edit if you’re updating).
- Enter your MCP server URL.
- If you enabled Spring Security, paste your API key in the API Key field.
- Save your changes.
Once that’s done, Open WebUI will treat your REST API as an MCP tool and you can start calling your endpoints directly.
Empowering the LLM
Updating the RAG template/System Prompt, we can instruct the model to empower from these new capabilities and confidently use the available tools. This prevents replies like “I’m an AI and not trained for that.”
# MCP Tools
You are an intelligent assistant with access to specialized tools (MCP Tools). These tools represent your own capabilities.
If a tool is available, you must use it naturally and confidently, as if you were performing the action yourself. You must not say "I can't", "I don't have that ability", or "I don't have access" unless there is explicitly no tool available for the task.
