deodorizing-ai-text

Detect and remove AI unnaturalness from Japanese text while preserving meaning.

Updated Jan 17, 2026
One-click install
npx skills add https://github.com/inakam/dotfiles-raspberry --skill deodorizing-ai-text
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: deodorizing-ai-text
Source: https://github.com/inakam/dotfiles-raspberry/tree/main/dot_claude/skills/deodorizing-ai-text
Command: npx skills add https://github.com/inakam/dotfiles-raspberry --skill deodorizing-ai-text

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill detects and removes the characteristic "AI smell" in Japanese text—translation-like phrasing, repetitive polite endings, unnatural lexical choices, and flattened subjectivity—and rewrites content to sound natural, human, and context-appropriate while preserving the original meaning.

Core Features & Use Cases

  • Layered AI-smell detection: Analyzes text across visual/structural, syntactic/rhythmic, lexical/grammatical, semantic/pragmatic, and cultural-context layers to identify issues and prioritize fixes.
  • Deodorization and rewrite: Applies a dictionary of natural replacements, rhythm modulation, injection of humanizing noise and sensory details, and conversion of passive/nominal forms into active, conversational Japanese.
  • Reports & validation: Produces an analysis report and a deodorization completion report with before/after metrics and a checklist to verify rhythm, wording, subjectivity, and cultural fit.
  • Use Case: Post-edit AI-generated apology emails, product reviews, or blog drafts to add personal voice, concrete examples, and natural rhythm so the text reads like it was written by a native human.

Quick Start

Paste the Japanese text you want deodorized and ask "この文章を脱臭して、人間らしい自然な日本語に直して" in a single request.

Frequently Asked Questions about deodorizing-ai-text

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

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

AI smell in Japanese text refers to translation-like phrasing, repetitive polite endings, unnatural lexical choices, and flattened subjectivity. It is detected through layered analysis across visual, syntactic, lexical, semantic, and cultural-context layers to identify and prioritize unnaturalness.

How do I remove AI detection markers from a Japanese apology email?

Removing AI detection markers from a Japanese apology email involves applying deodorization-dictionary replacements and rhythm modulation. This injects humanizing noise and sensory details, converting passive forms into active, conversational Japanese while preserving the original meaning.

Can I post-edit AI-generated product reviews to add personal voice?

You can post-edit AI-generated product reviews to add personal voice by applying deodorization-dictionary replacements and rhythm modulation. This injects humanizing noise and sensory details, converting passive forms into active conversational Japanese while preserving the original meaning.

What is the best way to rewrite unnatural Japanese blog drafts?

The best way to rewrite unnatural Japanese blog drafts is through style-preserving rewriting that applies layered AI-smell detection. This varies rhythm, injects sensory details, and converts nominal forms into active Japanese, validated by a before-and-after metrics report.

What are the limitations of automated Japanese style editing for AI text?

A limitation of automated Japanese style editing is that it requires a complete text input to perform layered analysis and deodorization-dictionary replacements. Users must validate the output against a checklist to ensure rhythm, wording, subjectivity, and cultural fit meet human-like standards.