Add linkedin-automation skill

[red777777]

LinkedIn Automation Skill

Author: Andreas Kulpa (bigdataheaven.com)

⚠️ Disclaimer

This skill is for personal, non-commercial use only. Not intended for spam, mass outreach, or commercial automation services.

What it does

Full LinkedIn automation via Playwright with persistent browser session:

• Post — Create new posts (text + optional image)
• Comment — With proper @mention handling (typeahead dropdown interaction, progressive name matching)
• Edit/Delete comments — Find by text fragment
• Repost — With your own commentary
• Feed reading — Extract posts from your feed
• Analytics — Engagement stats for your posts (reactions, comments, reposts)
• Like Monitor — Detect when you like a post, suggest comments (for cron/heartbeat integration)
• Profile stats — Followers, profile views

Key features

• Resilient selector engine — Multi-strategy fallbacks (aria-label → placeholder → class → JS walk)
• Proper @mention handling — Types first name, waits for typeahead, progressively types last name, validates match before clicking
• Structured JSON output from all commands
• Debug screenshots on failure
• Content strategy & engagement guides in references/

Prerequisites

• Python 3.10+ with Playwright
• Logged-in LinkedIn browser session (persistent Chromium profile)

Structure

linkedin-automation/
├── SKILL.md
├── references/
│   ├── content-strategy.md
│   ├── dom-patterns.md
│   └── engagement.md
└── scripts/
    ├── linkedin.py (CLI entry point)
    ├── requirements.txt
    └── lib/
        ├── actions.py, analytics.py, browser.py
        ├── feed.py, like_monitor.py, profile.py
        └── selectors.py (resilient selector engine)

[openclaw-barnacle]

Thanks for the pull request! This repository is read-only and is automatically synced from https://clawhub.ai, so we can’t accept changes here. Please make updates on the website instead.

[red777777]

Update: Removed hardcoded personal paths and names from the skill code (commit c67abff):

• browser.py: replaced /home/akulpa/... with ~/ (expanduser)
• like_monitor.py: replaced hardcoded author name with dynamic detection from LinkedIn sidebar

No credentials or API keys were found in the skill files.