Run a LinkedIn Account with Eleven Skills and a Local Humanizer

Most LinkedIn automation either posts for you and risks your account, or writes posts that read like a model wrote them. The LinkedIn agent skill — eleven free Claude skills by Jake Schincariol, MIT-licensed, no signup, no API key — takes a third path: the skills write, you post, and a local humanizer scores every draft before you see it.

Why This Skill Matters

The eleven skills cover the whole surface of running an account: /li-post turns one idea into a post with three hook options drawn from 21 formulas, /li-comment writes nine types of comments on other people's posts, /li-reply sorts the thread under yours into lead / substance / peer / support / noise and answers in that order, /li-profile scores your profile against a 12-part rubric out of 100, and /li-plan lays out the week — what to post, when, and the 10 people to engage with. The rest handle document posts, repurposing one video, newsletter, or transcript into a week of standalone posts, DM sequences, inbox triage, and a post-mortem that ranks published posts by engagement rate and reach multiple rather than impressions.

The humanizer is the reason the rest are usable. /li-human ships two dependency-free Python scripts that run on your machine: humanize.py cleans a draft, detect.py scores it on five checks — burstiness, specificity, slop density, fingerprint, voice — with the mean weighted at 60% and the weakest single check at 40%, because a detector only needs one signal to fire.

Installation

Paste this into Claude:

https://github.com/Jakeschincariol/linkedin-agent-skill

Install this skill, then confirm /li-post works.

Or install it yourself in Claude Code:

git clone https://github.com/Jakeschincariol/linkedin-agent-skill.git
cp -r linkedin-agent-skill/skills/li-* ~/.claude/skills/

Or as a plugin:

/plugin marketplace add Jakeschincariol/linkedin-agent-skill
/plugin install linkedin-agent

Project-local instead of global: copy the same folders into your repo's .claude/skills/. No Claude Code at all? Paste any single SKILL.md at the top of a chat and it runs as a mode — you lose the two Python tools, which is most of the point of /li-human, but the rest works.

Real Workflow: One Idea In, a Scored Draft Out

1. Spend ten minutes on templates/voice.md first. Copy it to ~/.claude/linkedin/voice.md and fill it in, or paste three of your own posts into Claude and say "write my voice.md from these". Every skill reads that file; the README's warning is that skipping it makes everything sound like everyone else.
2. Give /li-post one idea. It returns three hook options from the formula library and one full draft — humanized before you see it.
3. Prove the delta yourself:

python3 humanize.py draft.txt --report      # clean it, show every change
python3 detect.py draft.txt                  # score it, five checks
python3 detect.py before.txt after.txt       # prove the delta

4. Read the report. Structures like "It's not just X, it's Y", rule-of-three triads, one-word rhetorical questions, hashtag walls, reflex engagement bait, and uniform sentence length get flagged instead of auto-fixed — changing a sentence's shape needs judgment, so those come back to you for a rewrite. On the README's deliberately terrible example draft, the panel flags 24.8; after humanize.py it reaches 69.7, a REVIEW verdict — the last stretch to PASS is the part the script deliberately leaves to you.
5. Copy the final block and paste it into LinkedIn yourself.

Tips

• The lexicon is meant to be edited. skills/li-human/slop.json holds 113 stock words and phrases — delve, leverage, robust, "let that sink in" — with plain-English replacements, plus 17 invisible-character classes and 11 structural tells. Add your own pet peeves.
• Invisible characters are a real, checkable fingerprint: zero-width spaces and joiners, word joiners, soft hyphens, byte-order marks, Unicode tag characters, and non-breaking and narrow spaces. Your keyboard does not make these; they survive copy-paste and are invisible in every editor. The humanizer strips them automatically.
• Nothing gets fabricated under your name: if a draft needs a number you have not given, it comes back with {{your number}} in it and a flag, every time.
• /li-audit is the post-mortem for what you have already published — it ranks it by engagement rate and reach multiple, not impressions.

When Not to Use This

These skills do not post to LinkedIn, and the README is upfront about why: there is no official API for posting to a personal profile without an approved partner app, and automating the site with a browser or a third-party tool violates LinkedIn's User Agreement and gets accounts restricted. Every skill ends the same way — a copy-ready block, and you paste it. If you want scheduled auto-posting, this is not that tool. And the five checks are local heuristics modelled on the signals public detectors key on; they are not GPTZero, Originality, Copyleaks, Winston, or Turnitin, and they cannot promise those verdicts.

See the leaderboard for more skills.