linkedin-commander

LinkedIn CLI

Command-line tool for managing your LinkedIn account. Read your feed, view profiles, analyze post engagement, manage connections, search people/companies/jobs, and send messages — all from the terminal.

Uses Nodriver (CDP-based browser automation) under the hood — no WebDriver binary, no navigator.webdriver flag, undetectable by LinkedIn's PerimeterX/HUMAN anti-bot system.

---

What's in this Repo

This repo contains 3 things — install what you need:

What	For whom	Install
LinkedIn CLI	Everyone	Binary or pip install
CLI Reference Plugin	AI/Claude Code users who want CLI knowledge for their agents	/plugin install linkedin-cli-reference
LinkedIn Commander Plugin	Marketers who want a full AI marketing team	/plugin install linkedin-commander

1. LinkedIn CLI (the tool itself)

Terminal tool to interact with LinkedIn. Works standalone, no AI needed.

linkedin-cli profile show williamhgates
linkedin-cli posts analytics 7438407227096539136
linkedin-cli search people "software engineer" --limit 5

2. CLI Reference Plugin (for AI agents)

A single skill that teaches Claude Code (or any agent) how to use all CLI commands — correct syntax, flags, --json output shapes.

# In Claude Code:
/plugin marketplace add sabania/linkedin-cli
/plugin install linkedin-cli-reference

After installing, any agent or subagent can load the linkedin-cli-reference skill and knows how to call the CLI via Bash.

3. LinkedIn Commander Plugin (full marketing system)

Self-learning marketing system with 9 AI agents as a marketing team. Content strategy, post analysis, lead generation, competitor tracking, signal detection — all delta-based with human-in-the-loop.

# In Claude Code:
/plugin marketplace add sabania/linkedin-cli
/plugin install linkedin-commander
/setup

Full documentation: plugin/README.md

---

CLI Installation

Prerequisites

• Google Chrome installed (Nodriver communicates with it via CDP — no separate ChromeDriver needed)

Quick Install

Windows (PowerShell):

irm https://raw.githubusercontent.com/sabania/linkedin-cli/master/install.ps1 | iex

macOS / Linux:

curl -fsSL https://raw.githubusercontent.com/sabania/linkedin-cli/master/install.sh | bash

Installs to ~/.local/share/linkedin-cli (Unix) or %LOCALAPPDATA%\linkedin-cli (Windows) and automatically adds it to PATH. Restart terminal afterwards.

Manual Install

Download ZIP from Releases, extract and manually add directory to PATH:

OS	File
Windows	linkedin-cli-windows.zip
macOS	linkedin-cli-macos.tar.gz
Linux	linkedin-cli-linux.tar.gz

From Source (Developers)

git clone https://github.com/sabania/linkedin-cli.git
cd linkedin-cli
pip install -r requirements.txt
python main.py login

Uninstall

# Windows
& ([scriptblock]::Create((irm https://raw.githubusercontent.com/sabania/linkedin-cli/master/install.ps1))) -Uninstall

# macOS / Linux
curl -fsSL https://raw.githubusercontent.com/sabania/linkedin-cli/master/install.sh | bash -s -- --uninstall

---

Login

# Opens Chrome — you log in, CLI captures cookies automatically. 2FA works.
linkedin-cli login

# On headless servers (no display): paste li_at cookie instead
# Get it from: Browser → linkedin.com → DevTools (F12) → Application → Cookies → li_at
linkedin-cli login --cookie "AQE..."

# Verify
linkedin-cli whoami

Login only needs to be done once. Session cookies are stored at ~/.linkedin-cli/cookies.json.

---

CLI Commands

Settings

linkedin-cli config show                                # Show all settings + daily usage
linkedin-cli config set rate_limits.daily_limit 120     # Max API calls per day
linkedin-cli config set rate_limits.calls_per_minute 20 # Burst limit per minute
linkedin-cli config set browser.headless false          # Show browser window (debug)
linkedin-cli config reset                               # Reset to defaults

Profile

linkedin-cli profile show <user>            # View profile
linkedin-cli profile posts <user> -n 5      # Posts with engagement stats
linkedin-cli profile contact <user>         # Contact info
linkedin-cli profile skills <user>          # Skills
linkedin-cli profile experiences <urn>      # Work experience
linkedin-cli profile connections <urn>      # Connections
linkedin-cli profile views                  # Who viewed your profile
linkedin-cli profile network <user>         # Followers, connections count
linkedin-cli profile raw <user>             # Raw JSON dump

Feed

linkedin-cli feed list -n 10                # Your feed
linkedin-cli feed list --no-promoted        # Without sponsored posts

Posts