HR Assist MCP Server
An MCP server that automates HR onboarding tasks including employee management, email, ticketing, meeting scheduling, and leave management, enabling AI agents to handle new-hire workflows from a single prompt.
README
🧑💼 HR Assist — MCP Server
An MCP (Model Context Protocol) server that automates HR onboarding tasks — HRMS registration, welcome emails, IT ticketing, meeting scheduling, and leave management — so an AI agent like Claude can handle the entire new-hire workflow from a single prompt. 🚀
✨ Features
| Area | What it does |
|---|---|
| 👤 Employee Management | Add employees, fetch details by (fuzzy) name search, look up managers & direct reports |
| Send plain-text or HTML emails (with optional attachments) via SMTP | |
| 🎫 Ticketing | Raise, update, and list IT/equipment tickets (laptop, ID card, etc.) |
| 📅 Meetings | Schedule, list, and cancel meetings for any employee |
| 🌴 Leave Management | Check leave balance, apply for leave, view leave history |
| ⚡ Ready-made Prompt | Built-in onboard_new_employee MCP prompt that chains all of the above into one onboarding flow |
🗂️ Project Structure
atliq-hr-assist/
├── server.py # FastMCP server — defines all MCP tools & the onboarding prompt
├── emails.py # EmailSender — SMTP email utility (plain/HTML + attachments)
├── utils.py # seed_services() — seeds in-memory demo data
├── HRMS/
│ ├── __init__.py
│ ├── employee_manager.py # EmployeeManager — employee CRUD + org hierarchy
│ ├── meeting_manager.py # MeetingManager — schedule/list/cancel meetings
│ ├── leave_manager.py # LeaveManager — leave balance & history
│ ├── ticket_manager.py # TicketManager — IT/equipment tickets
│ └── schemas.py # Pydantic models for all entities
├── pyproject.toml
├── .python-version
├── .gitignore
└── .env # 🔒 local secrets — never committed
📌 Note:
server.pydoesfrom HRMS import *, so the manager classes (employee_manager.py,meeting_manager.py, etc.) andschemas.pyshould live inside anHRMS/package folder alongside__init__.py.
🛠️ Requirements
- 🐍 Python 3.10+
- Dependencies (see
pyproject.toml):fastmcp>=3.4.5mcp[cli]>=1.9.4pydantic>=2.13.4python-dotenv
⚙️ Setup
1. Clone & install
git clone https://github.com/Vraj-Data-Scientist/hr-onboarding-agent-with-MCP.git
cd hr-onboarding-agent-with-MCP
# using uv (recommended)
uv sync
# or using pip
pip install -e .
2. Configure environment variables
Create a .env file in the project root (already git-ignored 🔒):
EMAIL=your-smtp-email@gmail.com
EMAIL_PWD=your-app-specific-password
⚠️ Security note: Use a Gmail App Password, not your real account password, and never commit
.envto version control. If real credentials were ever pushed to a repo, rotate/revoke them immediately.
3. Run the server standalone (optional, for testing)
python server.py
This starts the MCP server over stdio transport.
🔌 Connecting to Claude as an MCP Client
This server speaks MCP over stdio, so it plugs directly into any MCP-compatible Claude client.
🖥️ Option A — Claude Desktop
- Open your Claude Desktop config file:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
- macOS:
- Add this server under
mcpServers:
{
"mcpServers": {
"hr-onboarding-agent": {
"command": "uv",
"args": [
"run",
"--directory",
"/absolute/path/to/hr-onboarding-agent-with-MCP",
"mcp",
"run",
"server.py"
],
"env": {
"EMAIL": "your-smtp-email@gmail.com",
"EMAIL_PWD": "your-app-specific-password"
}
}
}
}
- Restart Claude Desktop. A 🔨 tools icon should appear, showing the
hr-assistserver connected with tools likeadd_employee,send_email,create_ticket,schedule_meeting, and more. - Prompt Claude naturally, e.g.:
"Onboard a new employee named Priya Nair reporting to Tony Sharma."
💡 Using the built-in onboarding prompt
The server exposes a ready-made MCP prompt, onboard_new_employee, which Claude Desktop (via the / prompt picker) and Claude Code can surface directly — just supply employee_name and manager_name, and the full onboarding flow runs end-to-end.
🧪 Example
Prompt:
Onboard a new employee with the following details:
- Name: Vraj Dobariya
- Manager Name: Tony Sharma
What happens automatically:
- 👤 Employee added to HRMS with correct manager mapping
- 📧 Welcome email sent with generated login
- 📨 Manager notified of the new hire
- 🎫 Tickets raised for laptop, ID card, and other equipment
- 📅 Introductory meeting scheduled between employee and manager
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。