humankit

Rewrite prose to remove AI-generated writing patterns while preserving all original facts.

1|Updated Jul 12, 2026
One-click install
npx skills add https://github.com/mimukit/skills --skill humankit-mimukit
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: humankit
Source: https://github.com/mimukit/skills/tree/main/skills/humankit
Command: npx skills add https://github.com/mimukit/skills --skill humankit-mimukit

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? AI-generated text carries recognizable tells: em-dash overuse, puffery, filler phrases, rule-of-three structures, and vocabulary like "delve" and "tapestry." This Skill rewrites that prose so it reads like a specific human wrote it, without inventing facts or flattening the content into a summary. ## Core Features & Use Cases - Tell detection and removal: Scans for a catalog of AI writing patterns (inflated significance, promotional tone, AI vocabulary, false ranges, chatbot residue, formatting tells) plus two falsifiable cut tests, then rewrites around them. - Fact-safe rewriting: Never introduces a fact, name, number, date, quote, or citation absent from the source; concreteness comes from the source or the text stays plain. - Voice matching: When given a sample of the user's own writing, it matches that sample's sentence length, vocabulary, and punctuation habits instead of a generic style. - Use Case: Paste a ChatGPT-drafted blog paragraph and ask to "de-slop this" to receive a rewritten version, a note on what still read as AI, and a one-line summary of changes. ## Quick Start Paste your draft into the chat and ask the agent to humanize it and remove the AI writing tells.

Frequently Asked Questions about humankit

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I make AI-generated text sound human?

Paste the text and ask to humanize it or remove the AI tells. The skill scans for patterns like em-dash overuse, promotional adjectives, and filler phrases, then rewrites the prose with varied rhythm and plain constructions while keeping every original fact.

How to remove AI writing tells from a draft without rewriting it?

Ask for a review instead of a rewrite. The skill runs only the detection pass and reports the located tells with line references, so you can decide which patterns to fix yourself before re-running the rewrite.

Can it match my own writing style when humanizing text?

Yes. Supply a sample of your own writing and the skill matches its sentence length, vocabulary level, punctuation habits, and transitions. The sample overrides every style rule, including the em-dash ban.

Does humanizing text change the facts or meaning?

No. The rewrite never adds a fact, name, number, date, quote, or citation that is not in the source or supplied by you. Where the source lacks concrete detail, the text is cut to the plain version rather than filled with invented specifics.

When should I not use AI-text humanizing on a document?

Do not use it on instruction files for models such as SKILL.md, CLAUDE.md, system prompts, or rules files, since the flagged patterns are functional levers there. It is also not meant to disguise machine-written work where disclosure is required, such as academic submissions.