Logtime Summarizer

Table of Contens

• 1. Overview

• 2. Key Features

• 3. Tech Stacks

• 4. Installation

  • Getting Started

  • Installing Python

  • Installing uv package manager

  • Setup dependecies

  • Setup .env configuration

• 5. Test Server

  • Using npx as inspector

  • Using uv

• 6. Future Roadmap

Related MCP server: TrackingTime MCP

Overview

This Chatbot power with Model Context Protocol (MCP) as standardized way to connect AI models to different data sources and tools.

Key Features

The main feature is a Chatbot Assistant that answers summary about:

• Projects: Check on budgets, timelines, and overall progress.

• Employees: View team performance, workload, and efficiency.

• Clients: Track project status and resources for each client account.

Tech Stacks

• Client: Streamlit, Cursor

• Server: Python, FastAPI, OpenAI, MCP

Installation

Getting started

Clone the project using HTTPS

  git clone https://git.gits.id/ai-for-gits/ai-multi-agent-crew-ai-be.git

Installing Python

Recommendation to use python version 3.12.10, you can get it here or directly download the .exe file by click this url

Installing uv package manager

This project power with UV Python package and project manager. Here common method for installat UV

  pip install uv

You can learn more other methode on UV Documentation

Setup dependecies

Initialization uv package manager by create virtual enviroment and install the dependencies

  uv init
  uv venv --python 3.12
  uv add -r requirements.txt

Your folder will update with .venv, pyproject.toml, main.py, and uv.lock

logtime-summarize
├── .venv
├── ...
├── api
│   └── ...   
├── front
│   └── ...
├── mcp-server
│   └── ...
├── main.py
├── pyproject.toml
├── README.md
├── requirement.txt
└── uv.lock

Setup .env configuration

Make file .env, do double enter while run the script below

  echo > ".env"

Copy setup on .env.example to .env

Setup credentials configuration

Make file credentials json file, do double enter while run the script below

  echo > "./mcp-server/credentials.json"

Copy setup on credentials.example.json to credentials.json

Test Server

Using npx as inspector

  npx @modelcontextprotocol/inspector    

• COMMAND: {add-your-own-path}/logtime-summarizer/mcp-server/main.py

• ARGUMENTS: {add-your-own-path}/logtime-summarizer/mcp-server/main.py

• CONFIGURATE:

  • ...

  • Inspector Proxy Address: copy from terminal after running npx

  • Proxy Session Token: copy from terminal after running npx

Got an error? Learn more about the inspector

Using uv

  uv run mcp dev ./mcp-server/main.py

do same things like npx exclude setup command and arguments

• CONFIGURATE:

  • ...

  • Inspector Proxy Address: copy from terminal after running uv

  • Proxy Session Token: copy from terminal after running uv

Configuration Host for MCP client (Cursor, Claude Desktop or other IDEs)

{
    "mcpServers": {
        "logtime-summarizer": {
            "command": "add-your-own-path}/.local/bin/uv.exe",
            "args": [
                "run",
                "--directory",
                "C:\\D\\Work\\daily-reminder\\utils\\gsheet",
                "stdio.py"
            ],
            "env": {
                "OPENAI_API_KEY": "<your-openai-api-key>"
            }
        }
    }
}

• Cursor ~/.cursor/mcp.json

  Learn more Cursor Model Contex Protocol (MCP)

• Trae ~/.cursor/mcp.json

  Learn more Trae Model Contex Protocol (MCP)

• Windsurf ~/.codeium/windsurf/mcp_config.json

  Learn more Windsurf Model Contex Protocol (MCP)

• Claude Desktop ~/Library/Application\ Support/Claude/claude_desktop_config.json

  Learn more Claude Desktop Model Contex Protocol (MCP)

• Claude Code ~/.claude.json

  Learn more Claude Code Model Contex Protocol (MCP)

Learn more other MCP Clients that support MCP

Future Roadmap

• Custom MCP Client (UI) using Streamlit

• Deploy public MCP server

• Integration MCP server with Slack Bot