li-human — Claude Code skill: what it does, how to install
li-human strips em dashes, stock phrases and invisible Unicode characters from a draft, then scores it on five checks. Useful before a post, comment or reply goes out.
MIT ★ 3,812 updated 2026-09-17
/plugin marketplace add Jakeschincariol/linkedin-agent-skill
/plugin install linkedin-agent@linkedin-agent-skillWhat li-human does
The skill runs a draft through two Python scripts. humanize.py cleans the text, detect.py grades it on five checks. Cleaning deletes invisible characters such as zero-width spaces, soft hyphens, BOMs and Unicode tag characters, normalises typography, and swaps stock phrases for plainer wording.
The word list sits in slop.json: over 100 stock words and phrases with replacements, 17 classes of invisible characters, 11 typographic substitutions, 11 structural tells. Edit the file freely. If it strips a word you use all the time, take that entry out yourself, or the word will vanish from every draft.
Structural tells are flagged, never rewritten: triads, staged reveals, one-word question lines, emoji, hashtag walls. You fix those lines by hand and run detect.py again. The verdict weights the average at 60% and the weakest check at 40%; PASS means 70+ with nothing below 55. Everything runs on your machine and nothing is uploaded, but these are local heuristics, not GPTZero or Turnitin, and their verdict is not something the skill can promise.
What it helps with
- remove zero-width characters from text pasted out of a chat window
- swap curly quotes and em dashes for plain typography
- find stock vocabulary such as
delveand replace it with simpler words - compare a draft with its cleaned version across five checks
- list the lines a script flagged but would not rewrite
Example prompts for the agent
- Clean
draft.txtwithhumanize.pyand show me the report - Strip the em dashes and stock phrases from this post, then rewrite the flagged lines yourself
- Score
before.txtandafter.txtwithdetect.pyand tell me which check is weakest
How to install li-human
1Plugin marketplace
/plugin marketplace add Jakeschincariol/linkedin-agent-skill/plugin install linkedin-agent@linkedin-agent-skillThis installs the whole linkedin-agent plugin with all its skills
2Skills CLI (Vercel)
npx skills add Jakeschincariol/linkedin-agent-skill --skill li-human -a claude-code3Manually: put the files in place
~/.claude/skills/li-human/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/li-human/
1Skills CLI (Vercel)
npx skills add Jakeschincariol/linkedin-agent-skill --skill li-human -a codex2Manually: put the files in place
.agents/skills/li-human/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add Jakeschincariol/linkedin-agent-skill --skill li-human -a cursor2Manually: put the files in place
.cursor/skills/li-human/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install Jakeschincariol/linkedin-agent-skill skills/li-human2Manually: put the files in place
.github/skills/li-human/SKILL.md1Skills CLI (Vercel)
npx skills add Jakeschincariol/linkedin-agent-skill --skill li-human -a gemini-cli2Platform CLI
gemini skills install https://github.com/Jakeschincariol/linkedin-agent-skill --path skills/li-humanManually: put the files in place
.devin/skills/li-human/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about li-human
How do I install li-human in Claude Code?
Add the marketplace with /plugin marketplace add Jakeschincariol/linkedin-agent-skill, then install the plugin: /plugin install linkedin-agent@linkedin-agent-skill.
What does li-human do?
It clears a draft of invisible characters, machine typography and stock phrases, then scores five signals: sentence-length variation, specificity, slop density, character fingerprint and voice.
Does a PASS score mean my text will not be flagged?
No. These are five local heuristics. They do not call GPTZero, Originality, Copyleaks, Winston or Turnitin, and they cannot reproduce those verdicts.
Source text
SKILL.md · MIT
--- name: li-human description: >- Strip the machine fingerprint out of any draft - em dashes, AI slop words, invisible watermark characters - and score it against a five-check detection panel before it goes out. Use whenever text needs to sound human, when the user says humanize, "does this sound like AI", "remove the em dashes", "de-slop this", "will this get flagged", or before any LinkedIn post, comment, reply or DM is shown to the user. --- # li-human Two tools live in this folder and they both actually run. Use them. Do not eyeball this. ```bash python3 humanize.py draft.txt --report # clean it, show what changed python3 detect.py draft.txt # score it, five checks python3 detect.py before.txt after.txt # prove the delta ``` Both read `slop.json`, which is the lexicon: 100+ stock words and phrases with plain-English replacements, 17 invisible character classes, 11 typographic substitutions, and 11 structural tells. It is meant to be edited. If the user has a word they always use that the lexicon strips, remove it from the file. ## What gets fixed automatically **1. Invisible characters.** Zero-width spaces and joiners, word joiners, soft hyphens, byte-order marks, Unicode tag characters, non-breaking and narrow spaces. A keyboard does not produce these. They survive copy-paste, they are invisible in every editor, and they are the single most mechanical thing in generated text. `humanize.py` deletes every one, including any remaining Unicode format character it does not have a name for. **2. Typography.** Em dash to comma, en dash to hyphen, curly quotes to straight, ellipsis to three dots, bullet character to hyphen. The em dash pass is the one that matters: it collapses ` — ` to `, ` and then cleans up the double punctuation that leaves behind. **3. The slop lexicon.** delve, leverage, robust, seamless, crucial, tapestry, testament to, moreover, "in today's fast-paced world", "let that sink in" and the rest, each swapped for a plain word, with capitalisation preserved and URLs left untouched. ## What does NOT get fixed automatically Structural tells get **flagged, not rewritten**, because changing the shape of a sentence needs judgement: - "It's not just X, it's Y" and "not only X but also Y" - Rule-of-three triads - Rhetorical one-word question lines: "The result?" - Rocket, fire, bulb, sparkle and dart emoji - Hashtag walls - Reflex engagement bait: "Thoughts?", "Agree?", "Who else?" - Uniform sentence length and uniform bullet length That list is your job. Rewrite each flagged line by hand, keeping the meaning, then re-run `detect.py`. This is the part that moves the score from REVIEW to PASS, and it is the part a script cannot do. ## The five checks `detect.py` scores five signals 0-100, higher is more human: | check | what it measures | machine looks like | | --- | --- | --- | | BURSTINESS | sentence-length variation | every sentence the same length | | SPECIFICITY | numbers, names, concrete markers per 100 words | abstract nouns, no figures | | SLOP DENSITY | lexicon hits per 100 words | stock vocabulary | | FINGERPRINT | invisible chars, em dashes, curly quotes per 1k chars | typographically perfect | | VOICE | contractions, person, structural tells | no contractions, staged reveals | The verdict weights the mean at 60% and the **weakest single check** at 40%, because a detector only needs one signal to fire. PASS needs an overall of 70+ with no check below 55. ## Say this honestly These are five local heuristics modelled on the signals public detectors key on. They run entirely on the user's machine and nothing is uploaded. They are **not** GPTZero, Originality, Copyleaks, Winston or Turnitin, they do not call those APIs, and they cannot promise those verdicts. Fixing what they measure does tend to move those numbers, because they are measuring the same underlying things. That is the claim. Do not make a bigger one on the user's behalf, and do not tell a user their text is undetectable. ## Order of operations 1. `humanize.py draft.txt -o clean.txt --report` 2. Read the structural flags. Rewrite those lines yourself. 3. `detect.py draft.txt clean.txt` to show the before and after. 4. If the verdict is not PASS, fix the weakest check named in the output and go again. Two rounds is normal. Five means the draft was written by formula, and the fix is a different draft, not more passes. 5. Show the user the cleaned text and the score. Never the score alone.
© Jake Schincariol · MIT · Jakeschincariol/linkedin-agent-skill