LinkedIn Job Sniper is an automated AI agent that hunts for job opportunities on LinkedIn (and other platforms), evaluates them against your resume using LLMs (GPT-4o, DeepSeek, etc.), and sends you a daily summary of the top matches directly to your inbox.

Stop doom-scrolling LinkedIn. Let the AI agent work for you 24/7.

• 🕵️ Automated Scraping: Scrapes the latest jobs from LinkedIn using python-jobspy (with Proxy support to avoid bans).
• 🧠 AI-Powered Analysis: Uses LangChain and LLMs (OpenAI/DeepSeek) to semantically score jobs based on your actual resume, not just keywords.
• 📧 Daily Email Digest: Sends a beautifully formatted HTML email with the Top 10 matched jobs, including an AI-generated "Reason for Match".
• ☁️ Zero Server Cost: Runs entirely on GitHub Actions (Free Tier).
• 🛡️ Robust & Resilient: Built-in retry logic, proxy rotation, and error handling for stable operation.

1. Trigger: Runs automatically every day (via Cron) or manually via GitHub Actions.
2. Scrape: Fetches job listings based on your search criteria (e.g., "Software Engineer" in "Singapore").
3. Evaluate: The AI Agent reads your resume and creates a relevance score (0-100) and a reason for every job found.
4. Report: Filters the best jobs and emails them to you.

Before starting, you need to prepare the following accounts and information:

1. Open your browser and go to https://github.com
2. Click "Sign up" in the top right corner
3. Enter:
   • Email address
   • Password (at least 15 characters, or 8 characters with numbers and lowercase letters)
   • Username (choose any, e.g., job-hunter-2026)
4. Verify your email (check inbox and click verification link)
5. Log in to GitHub

1. Visit the original project: https://github.com/tao-991/LinkedInJobSniper
2. Click the "Fork" button in the top right corner (fork icon)
3. On the popup page, click the green "Create fork" button
4. Wait a few seconds, the page will redirect to YourUsername/LinkedInJobSniper
5. ✅ Confirm the top left shows YourUsername/LinkedInJobSniper

Note: Forking means copying someone else's project completely to your account. All future operations will be on your own copy.

⚠️ Important: You cannot use your regular Gmail login password - you must generate a dedicated app password!

1. Go to https://myaccount.google.com/security
2. Scroll down to find the "Signing in to Google" section
3. Click "2-Step Verification"
4. Follow the prompts to set it up (usually using your phone number to receive verification codes)

1. Return to https://myaccount.google.com/security
2. In the search bar, type "App passwords"
3. Click "App passwords" in the search results
4. You may need to enter your Gmail password again
5. Under "Select app" dropdown, choose "Mail"
6. Under "Select device" dropdown, choose "Other"
7. Enter a name, e.g., JobSniper
8. Click "Generate"
9. 🔑 Copy the 16-character password (format like abcd efgh ijkl mnop)
10. ✅ Save this password in Notepad - you'll need it later

Tip: This password is only shown once. If you forget it, you'll need to generate a new one.

You need to choose an AI service provider. DeepSeek (cheap) or OpenAI (stable) are recommended.

1. Visit https://platform.deepseek.com
2. Click "Sign Up" or "Log In" in the top right
3. After registration/login, click "API Keys"
4. Click "Create API Key"
5. Give the key a name, e.g., JobSniper
6. 🔑 Copy the generated key (format like sk-xxxxxxxxxxxx)
7. Save to Notepad
8. Add $5-10 USD credit (will last a long time)

Additional Configuration:

• API Base URL: https://api.deepseek.com/v1
• Also save to Notepad

1. Visit https://platform.openai.com
2. Click "Sign up" or "Log in"
3. After logging in, click "API keys" in the left menu
4. Click "Create new secret key"
5. Give the key a name, e.g., JobSniper
6. 🔑 Copy the generated key (format like sk-proj-xxxxxxxxxxxx)
7. Save to Notepad
8. Add $5-10 USD credit

Additional Configuration:

• API Base URL: Not needed (leave empty)

⚠️ Why do you need a proxy? LinkedIn blocks bot access. A proxy disguises your requests as coming from real users.

1. IPRoyal (https://iproyal.com/residential-proxies/)

   • Price: $7 = 1GB traffic (lasts several months)

2. Smartproxy (https://smartproxy.com)

   • Price: $8.5 = 1GB traffic

3. ThorData (https://thordata.com)

   • Price: $1.2 = 1GB traffic

1. Visit https://iproyal.com/residential-proxies/

2. Click "Get Started"

3. Register an account and log in

4. Select "Pay As You Go" plan

5. Purchase 1GB traffic (about $7)

6. Go to Dashboard

7. Click "Residential Proxies"

8. Click "Generate Proxy"

9. You'll see proxy information in a format like:

   Host: proxy.iproyal.com
   Port: 12321
   Username: your_username
   Password: your_password
   

10. 🔑 Assemble into complete URL (very important - format must be exact):

    http://username:password@host:port
    

    Real Example:

    http://john123:pass456@proxy.iproyal.com:12321
    

11. Save to Notepad

⚠️ Format Notes:

• Must start with http:// (not https://)
• Colon and @ symbol positions must be correct
• No spaces allowed
• If password contains special characters like @ #, they need to be URL-encoded

1. Open your resume (Word, PDF, or other format)
2. Select all text and copy (Ctrl+C or Command+C)
3. Paste into Notepad
4. Remove all formatting, keep only plain text
5. Ensure it includes:
   • Your skills
   • Work experience
   • Education
   • Projects
6. Save - you'll need this later

Example Resume Format:

Name: John Smith
Email: john.smith@gmail.com
Skills: Python, JavaScript, React, Node.js
Work Experience:
- ABC Company, Software Engineer, 2020-2023
  - Developed XYZ system, improved efficiency by 30%
Education:
- MIT, Computer Science BS, 2016-2020

Now tell GitHub all your keys and configuration so the program can run automatically.

1. Make sure you're on your forked project page (URL should be https://github.com/YourUsername/LinkedInJobSniper)
2. Click the "Settings" tab in the top navigation
3. In the left menu, find "Secrets and variables"
4. Click to expand, select "Actions"

Now you need to add 7 secrets. For each secret, follow these steps:

1. Click the green "New repository secret" button
2. In the "Name" field, enter the secret name (see table below)
3. In the "Secret" field, paste the corresponding content
4. Click "Add secret"

7 Secrets to Add:

After completion, you should see 7-10 secrets on the Secrets page:

• ✅ API_KEY
• ✅ API_BASE
• ✅ EMAIL_SENDER
• ✅ EMAIL_PASSWORD
• ✅ EMAIL_RECEIVER
• ✅ PROXY_URL
• ✅ RESUME_TEXT (if used)
• ✅ CRITERIA (if used)
• ✅ GCP_CREDENTIALS_JSON (if used)
• ✅ RESUME_FILE_ID (if used)

If you want to modify the job type and location being searched, you can edit the code file.

1. On your project homepage, click the "main.py" file

2. Click the ✏️ pencil icon in the top right (Edit this file)

3. Find these lines of code (around lines 20-30):

   SEARCH_TERM = "Software Engineer"
   LOCATION = ["Singapore"]
   RESULTS_WANTED = 20

4. Modify to what you want:

   SEARCH_TERM = "Product Manager"  # Change to your desired position
   LOCATION = ["New York"]  # Change to your desired city
   RESULTS_WANTED = 30  # Number of jobs to search per run

5. Click the green "Commit changes" button in the top right

6. In the popup, click "Commit changes" again

1. On the project homepage, click the "Actions" tab at the top
2. If you see a yellow notice saying "Workflows aren't being run...", click the green button "I understand my workflows, go ahead and enable them"
3. In the left menu, click "LinkedIn Job Sniper"
4. Click the "Enable workflow" button on the right (if present)

Before waiting for the scheduled task, run it manually once to ensure configuration is correct.

1. Make sure you're on the "Actions" tab
2. In the left sidebar, click "LinkedIn Job Sniper" workflow
3. On the right, click the "Run workflow" dropdown
4. Keep Branch: main unchanged
5. Click the green "Run workflow" button
6. Refresh the page - you'll see a yellow spinning circle
7. Wait 3-5 minutes, the yellow circle will turn into:
   • ✅ Green checkmark: Success! Check your email
   • ❌ Red X: Failed - need to check configuration

1. Click the failed run record
2. Click the "run-job-sniper" task
3. Expand the red step to view error messages
4. Common errors:
   • Authentication failed: Gmail App Password is wrong - regenerate it
   • Proxy error: Proxy URL format is wrong or proxy service is down
   • API key invalid: AI API key is wrong or insufficient balance
   • 403 Forbidden: Proxy IP is blocked - try a different proxy

The project is already configured to run automatically every day at 8 AM Beijing time by default. If you want to change the time:

1. Click the ".github/workflows" folder on the project homepage

2. Click the "linkedin-job-sniper.yml" file

3. Click the ✏️ pencil icon to edit

4. Find this line:

   - cron: '0 0 * * *'  # Runs at 00:00 UTC daily

5. Modify to your desired time (uses UTC timezone, 8 hours behind Beijing time)

   • For example, to run at 8 AM Beijing time daily, change to:

     - cron: '0 0 * * *'  # UTC 0:00 = Beijing 08:00

   • To run at 10 PM Beijing time daily, change to:

     - cron: '0 14 * * *'  # UTC 14:00 = Beijing 22:00

6. Click "Commit changes" to save

Time Conversion Tip: Beijing Time - 8 hours = UTC Time
Example: Beijing 20:00 - 8 = UTC 12:00

Common UTC Times:

• 0 0 * * * = 8:00 AM Beijing / 12:00 AM UTC
• 0 6 * * * = 2:00 PM Beijing / 6:00 AM UTC
• 0 12 * * * = 8:00 PM Beijing / 12:00 PM UTC
• 0 14 * * * = 10:00 PM Beijing / 2:00 PM UTC

If you want to test it on your local machine:

1. Clone the repository

   git clone https://github.com/tao-991/LinkedInJobSniper.git
   cd LinkedInJobSniper

2. Install Dependencies

   pip install -r requirements.txt

3. Setup Environment Variables Create a .env file in the root directory:

   OPENAI_API_KEY=sk-xxxx
   API_BASE=https://api.deepseek.com/v1  # Optional
   EMAIL_SENDER=me@gmail.com
   EMAIL_PASSWORD=xxxx
   EMAIL_RECEIVER=me@gmail.com
   PROXY_URL=http://user:pass@host:port
   RESUME_TEXT="Paste your resume text here or use resumes/ folder"

4. Add Resume (Optional) Place your resume PDF in a folder named resumes/ (e.g., resumes/my_cv.pdf). The script will prioritize the file over the environment variable.

5. Run

   python main.py

You can modify main.py to change search parameters:

SEARCH_TERM = "Software Engineer"
LOCATION = ["Singapore"]
RESULTS_WANTED = 20  # Number of jobs to scrape per run

• Initial Release: LinkedIn Job Sniper with Gmail integration.
• Add support for users to enter their job hunting preferences. AI will also consider these preferences when scoring jobs.
• Add support for allowing users to upload their resume files directly.
• Integrate with more job platforms (Indeed, Glassdoor, etc.).
• Add support to more email providers (Outlook, Yahoo, etc.).

This tool is for educational purposes only. Web scraping may violate LinkedIn's User Agreement. Use this tool responsibly and at your own risk. The author is not responsible for any account bans or limitations.

Tao

• Powered by Python, LangChain, and Coffee ☕.