LinkedIn Easy Apply – Standalone Automation

A standalone Python + Selenium project that logs into LinkedIn, searches jobs with configurable keywords and location, and applies only to Easy Apply jobs. Applications are tracked to a CSV or JSON file, with configurable rate limiting to reduce detection risk.

Features

• Login via environment variables or .env (no passwords in code)
• Configurable job search: keywords, location, remote filter
• Easy Apply only: detects Easy Apply jobs, fills saved answers where possible, submits; skips or flags multi-step/complex applications
• Application tracking: job title, company name, job URL, date applied (CSV or JSON)
• Rate limiting: configurable delays between actions and between applications
• Single config: config.json and/or env vars for keywords, location, delays, optional resume path

Requirements

• Python 3.10+
• Chrome browser (for Selenium WebDriver)

Setup

1. Clone and create virtualenv

cd linkedin-easy-apply
python -m venv venv

Windows (PowerShell):

.\venv\Scripts\Activate.ps1

Windows (cmd) / macOS / Linux:

# Windows cmd
venv\Scripts\activate.bat
# macOS / Linux
source venv/bin/activate

2. Install dependencies

pip install -r requirements.txt

3. Configure credentials (required)

Create a .env file in the project root (never commit this file):

cp .env.example .env

Edit .env and set:

• LINKEDIN_EMAIL – your LinkedIn email
• LINKEDIN_PASSWORD – your LinkedIn password

4. Optional: config file

Copy the example config and edit as needed:

cp config.example.json config.json

You can override any of these with environment variables (see below).

Configuration

Environment variables

Variable	Description
LINKEDIN_EMAIL	LinkedIn login email (required)
LINKEDIN_PASSWORD	LinkedIn password (required)
CONFIG_FILE	Path to JSON config (default: config.json)
LINKEDIN_KEYWORDS	Job search keywords / role (e.g. software engineer)
LINKEDIN_LOCATION	Location (e.g. United Kingdom)
LINKEDIN_WORK_TYPE	Work type: 1 On-site, 2 Remote, 3 Hybrid (or Remote/Hybrid/On-site)
LINKEDIN_JOB_TYPE	Job type: F Full-time, P Part-time, C Contract, T Temporary, V Volunteer, I Internship, O Other
LINKEDIN_DATE_POSTED	Date posted: r86400 24h, r604800 week, r2592000 month
LINKEDIN_EXPERIENCE_LEVEL	Experience: 1–6 (comma-separated), see below
LINKEDIN_FEW_APPLICANTS	Fewer than 10 applicants: true/false
LINKEDIN_GEO_ID	LinkedIn location ID (optional; overrides location if set)
DELAY_ACTIONS_SEC	Delay between actions in seconds
DELAY_APPLICATIONS_SEC	Delay after each application in seconds
RESUME_PATH	Optional path to resume/CV file for upload
TRACKING_FILE	Output file for applications (e.g. applications.json or applications.csv)
TRACKING_FORMAT	json or csv
MAX_APPLICATIONS	Max number of applications per run (e.g. 5 for testing; 0 = no limit)

Where to set role, start date, and filters

What you want	Where to set it	Config key (in search or env)
Role / job title	Search keywords	keywords / LINKEDIN_KEYWORDS
Location	Search	location / LINKEDIN_LOCATION
Work type (On-site / Remote / Hybrid)	Search	work_type / LINKEDIN_WORK_TYPE (1/2/3 or "On-site"/"Remote"/"Hybrid")
Job type (Full-time, Part-time, Contract, etc.)	Search	job_type / LINKEDIN_JOB_TYPE (F, P, C, T, V, I, O)
Date posted	Search	date_posted / LINKEDIN_DATE_POSTED (r86400, r604800, r2592000)
Experience level	Search	experience_level / LINKEDIN_EXPERIENCE_LEVEL (1–6, comma-separated)
Few applicants (<10)	Search	few_applicants / LINKEDIN_FEW_APPLICANTS (true/false)
Location by ID	Search	geo_id / LINKEDIN_GEO_ID (optional)
Start date (when you can start)	Easy Apply form answer	saved_answers.start_date
First name, last name, email, phone, city, cover letter, salary, sponsorship	Easy Apply form answers	saved_answers in config.json

Search filters reference (all under config.json → search, or env):

Key	Values	Description
keywords	Any string	Role / job title search
location	e.g. "United Kingdom"	Geographic location
work_type	1 or "On-site", 2 or "Remote", 3 or "Hybrid"	Work arrangement
job_type	F Full-time, P Part-time, C Contract, T Temporary, V Volunteer, I Internship, O Other	Job type
date_posted	r86400 (24h), r604800 (week), r2592000 (month)	When job was posted
experience_level	1 Intern, 2 Associate, 3 Junior, 4 Mid-Senior, 5 Director, 6 Executive (use "3,4" for multiple)	Experience level
few_applicants	true / false	Only jobs with fewer than 10 applicants
geo_id	LinkedIn geo ID	Optional; overrides location if set

Not available as URL filters (LinkedIn does not expose these in the search URL): salary range, industry, job function, benefits, sponsorship. You can add keywords like "visa sponsorship" in keywords, or apply those filters once manually in LinkedIn’s job search UI and then run the script.

Form answers (first name, last name, email, phone, city, start date, cover letter, salary, sponsorship, etc.) go in saved_answers and are used when filling the Easy Apply modal.

Config file (config.json)

Example (all optional if using env vars):

{
  "max_applications": 5,
  "search": {
    "keywords": "software engineer",
    "location": "United Kingdom",
    "work_type": "2",
    "job_type": "F",
    "date_posted": "r604800",
    "experience_level": "3,4",
    "few_applicants": false,
    "geo_id": ""
  },
  "rate_limiting": {
    "delay_between_actions_sec": 2,
    "delay_between_applications_sec": 30
  },
  "resume_path": "/path/to/resume.pdf",
  "tracking": {
    "output_file": "applications.json",
    "format": "json"
  },
  "saved_answers": {
    "first_name": "Your First Name",
    "last_name": "Your Last Name",
    "email": "your.email@example.com",
    "phone": "+44 ...",
    "city": "London",
    "cover_letter": "...",
    "salary": "",
    "sponsorship": "No",
    "start_date": "Immediately"
  }
}

Environment variables override config file values. Credentials must come from the environment (or .env), not from the config file.

How to run and test

Quick test (no applications submitted)

1. Setup once

   cd linkedin-easy-apply
   python -m venv venv
   .\venv\Scripts\Activate.ps1   # Windows PowerShell
   pip install -r requirements.txt

2. Add credentials
   • Copy .env.example to .env and set LINKEDIN_EMAIL and LINKEDIN_PASSWORD.
3. Dry run (login + search only; no Easy Apply clicks, no submissions):

   python main.py --dry-run

   You should see a Chrome window log in, open the job search, and then:
   Dry run: found N job cards, M with Easy Apply. Login and search OK. Run without --dry-run to apply.

Full run (apply to jobs)

1. Activate the virtualenv and ensure .env is set (see above).
2. Optional: copy config.example.json to config.json and set keywords, location, delays, saved_answers, and resume_path.
3. Run:

   python main.py

The script will:

1. Log into LinkedIn
2. Open job search with your keywords/location (Easy Apply filter applied)
3. Iterate over Easy Apply job cards: open each job, click Easy Apply, fill saved answers where possible, submit if it’s a single-step flow
4. Skip jobs already in the tracking file, multi-step applications, or when it can’t submit
5. Append each applied job to applications.json (or your configured CSV/JSON file)

Output (tracking)

• JSON (default): array of { "job_title", "company_name", "job_url", "date_applied" }.
• CSV: same fields as columns.

File path and format are set by TRACKING_FILE and TRACKING_FORMAT (or config.json).

Rate limiting

• delay_between_actions_sec: pause between normal actions (clicks, page loads).
• delay_between_applications_sec: longer pause after each successful application.

Increase these if you want to lower the risk of detection (e.g. 3–5 s for actions, 45–60 s between applications).

Security and compliance

• Do not put LinkedIn credentials in code or in config.json.
• Use .env for local secrets; .env is in .gitignore.
• Respect LinkedIn’s Terms of Service and use automation at your own risk. Prefer conservative delays and occasional use.

Project structure

linkedin-easy-apply/
├── .env.example       # Example env (copy to .env)
├── .gitignore         # Includes .env
├── config.example.json
├── config.py          # Loads config from env + optional config.json
├── linkedin_automation.py  # Login, search, Easy Apply (Selenium)
├── main.py            # Entry point, rate limiting, tracking loop
├── tracker.py         # CSV/JSON application tracking
├── requirements.txt
└── README.md

Troubleshooting

• Login fails: Check email/password in .env; ensure no 2FA or use an app password if required.
• No jobs / wrong jobs: Adjust keywords and location in config or env; search uses LinkedIn’s built-in Easy Apply filter.
• Applications not submitted: Many Easy Apply forms have multiple steps or custom questions; the script submits only when it finds a single-step Submit. Others are skipped.
• LinkedIn blocks or CAPTCHA: Increase delays and run less frequently.