LinkedIn Agent Skill
Eleven free LinkedIn skills — posts from 21 hook formulas, comments, profile scoring, and an AI humanizer.
About
LinkedIn Agent Skill by Jakeschincariol is an MIT-licensed package of eleven Claude skills for running a LinkedIn account, with 1,948 GitHub stars. One skill writes posts off 21 hook formulas, others handle comments, replies under your posts, profile scoring out of 100 with rewrites of what lost points, and weekly planning of what to post, when, and who to engage with. A humanizer strips em dashes, slop vocabulary, and invisible watermark characters, then scores the draft against a five-check detection panel before you see it. Nothing gets posted until you say yes — the skills write, you post. A voice.md template learns your writing style from your own past posts.
Worth knowing: spend ten minutes filling in the voice template first, or everything comes out sounding like everyone else. No-Claude-Code fallback exists — pasting any single SKILL.md into a chat runs it as a mode, minus the two Python tools that power the humanizer.
What it can do
- Post writer with 21 hook formulas
- Comment and reply drafting
- Profile scoring and rewriting
- Weekly content planning
- Humanizer with five-check detection panel
- Voice template for brand consistency
Setup
git clone https://github.com/Jakeschincariol/linkedin-agent-skill.git && cp -r linkedin-agent-skill/skills/li-* ~/.claude/skills/. Or: /plugin marketplace add Jakeschincariol/linkedin-agent-skill, then /plugin install linkedin-agent.
Prerequisites
Claude Code; fill in templates/voice.md with your own posts for best results.
Use cases
Works with
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Live GitHub repo: 1,948 stars, MIT license, not a fork, verified this session; 11 skills incl. humanizer with five-check detection panel, verbatim install commands from README; voice.md template for brand consistency noted as a prerequisite.
★ 1,948 GitHub stars
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.