# Naukri MCP Server

An MCP-based Naukri job search assistant that scrapes recent job
postings, processes them, tracks previously seen jobs, and exports
new opportunities to Excel with direct Naukri application links.

## Features

- Naukri job scraping
- Multiple roles
- Multiple locations
- Freshness filtering
- Experience filtering
- Duplicate detection
- Previously seen job tracking
- Application status tracking
- Match score
- Priority
- Excel export
- Direct Naukri application links
- Claude Desktop MCP integration

## Requirements

- Windows / macOS / Linux
- Python 3.10+
- Git
- uv
- Playwright
- Chromium
- Claude Desktop

## Installation

### 1. Clone

git clone https://github.com/pcsuper50/Naukri-mcp-server.git

cd Naukri-mcp-server

### 2. Install dependencies

uv venv

uv pip install -r requirements.txt

### 3. Install Playwright browser

uv run playwright install chromium

### 4. Create configuration

Copy:

config.example.py

to:

config.py

Update your:

- name
- experience
- skills
- target roles
- preferred locations
- excluded roles

## Test the scraper

uv run python scraper.py --roles "GenAI Engineer" --locations Pune --days 3 --max-pages 1

If the scraper is working you should see:

CARDS|20

and matched jobs.

## Full search

uv run python scraper.py --roles "GenAI Engineer" "AI ML Engineer" "LLM Engineer" "Machine Learning Engineer" "RAG Engineer" "Production Support" "L2 Support" "Application Support" --locations "Pune" "Bangalore" "Hyderabad" "Noida" "Gurgaon" "Remote" "Indore" "Ahmadabad" --days 3 --max-pages 5 --min-exp 1 --max-exp 6

## Claude Desktop Setup

Open the Claude Desktop configuration file.

Use content:

claude_desktop_config.example.json

as the template.

Replace:

C:\\PATH\\TO\\Naukri-mcp-server

with your actual project path.

Save the configuration and restart Claude Desktop.

## Daily Workflow

### Step 1

Run the scraper from VS Code[Fill your experience and role accordingly]:

uv run python scraper.py --roles "GenAI Engineer" "AI ML Engineer" "LLM Engineer" "Machine Learning Engineer" "RAG Engineer" "Production Support" "L2 Support" "Application Support" --locations "Pune" "Bangalore" "Hyderabad" "Noida" "Gurgaon" "Remote" "Indore" "Ahmadabad" --days 3 --max-pages 5 --min-exp 1 --max-exp 6

Wait until scraping completes.

### Step 2

Open:

CLAUDE_PROMPT.md

Copy the prompt into Claude Desktop.

Claude will:

1. Process jobs
2. Calculate statistics
3. Export new jobs to Excel
4. Retrieve new jobs
5. Display direct application links

## Output

Generated files are stored locally inside:

naukri/

Examples:

naukri_latest_jobs.xlsx
processed_jobs.json

These files are intentionally excluded from Git.

## Troubleshooting

### No jobs found

Run a smaller test:

uv run python scraper.py --roles "GenAI Engineer" --locations Pune --days 3 --max-pages 1

### Claude cannot connect

Check:

1. uv is installed
2. project path is correct
3. Claude Desktop config is valid
4. Claude Desktop was restarted
5. Chromium was installed

## License

MIT