sergebulaev/linkedin-skills — reverse-engineered prompt

Reverse engineered prompt

GitHub

Build me an open source Claude skills pack that helps people grow on LinkedIn without needing to write everything themselves.

I want users to be able to ask Claude for human sounding LinkedIn posts, thoughtful comments, replies, profile rewrites, draft audits, hook breakdowns, a seven day content plan, and a simple employee advocacy plan. It should remove obvious AI tells, make the writing feel more natural, and always show a draft first. Nothing should ever get posted without the user approving it.

Include optional setup for reading LinkedIn post text, comments, replies, and engagers through Apify if the user adds a token. Also include optional publishing or scheduling through Publora if they add their API key and LinkedIn platform ID. If those are missing, just ask the user to paste the content or copy the final draft manually.

Make it easy to install in Claude Code, Claude Desktop, claude.ai, or OpenClaw, with clear README instructions and simple Python helper files. Look up current docs online if needed.

Want more depth? Deep Reverse