skills.rest
matsumuratk
matsumuratkCOMMUNITY
Explore all skills by this author →
GitHub

genshijin

Compresses Japanese AI responses into terse keyword-style output to reduce token usage.

1|Updated Jul 12, 2026
Search Tags:#japanese#concise output#context optimization#token reduction#response compression#prompt style
One-click install
npx skills add https://github.com/matsumuratk/elmo-aws --skill genshijin-matsumuratk
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: genshijin
Source: https://github.com/matsumuratk/elmo-aws/tree/main/.claude/skills/genshijin
Command: npx skills add https://github.com/matsumuratk/elmo-aws --skill genshijin-matsumuratk

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Long, polite Japanese AI responses consume excessive tokens and bury technical answers in filler phrases. This Skill strips honorifics, hedging, and redundant grammar from responses while preserving full technical accuracy, cutting token usage by roughly 75%. ## Core Features & Use Cases - Three Intensity Levels: Choose between 丁寧 (polite but concise), 通常 (default, telegraphic style), and 極限 (keyword-only with abbreviations and arrows). - Technical Accuracy Preservation: Code blocks, error messages, and technical terms remain unchanged; only filler language is removed. - Safety-Aware Behavior: Automatically switches back to normal Japanese for destructive operation confirmations, then returns to compressed mode. - Use Case: During a long debugging session, activate the mode to get answers like "認証ミドルウェア バグ。トークン期限チェック <→<=。" instead of multi-sentence polite explanations, stretching your context window further. ## Quick Start Ask the AI to enable primitive mode by saying "原始人モード" or "トークン節約して", then continue your technical conversation as usual.

Frequently Asked Questions about genshijin

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

FAQPage Schema
How do I reduce token usage in Japanese AI conversations?▼

Enable this compression mode by saying "原始人モード", "短く", or "トークン節約". The AI then strips honorifics, filler words, and redundant particles from responses while keeping all technical content, reducing token usage by roughly 75%.

How do I change the compression intensity level?▼

Use the command /genshijin followed by a level name: 丁寧 (polite but concise), 通常 (default telegraphic style), or 極限 (keyword-only with abbreviations). The chosen level persists until changed or the session ends.

Does response compression affect code blocks and error messages?▼

No. Code blocks, error messages, and technical terms are preserved exactly as-is. Only natural language filler such as honorifics, hedging expressions, and redundant particles is removed.

When does the compressed mode turn off automatically?▼

It temporarily switches to normal Japanese when confirming destructive operations like DROP TABLE, then immediately returns to compressed mode. Saying "原始人やめて" or "通常モード" disables it entirely.

What are the limitations of extreme compression mode?▼

The 極限 level ignores Japanese grammar entirely, using only keywords, abbreviations, and arrows. This can reduce readability for nuanced explanations, so the default 通常 level is recommended for most technical discussions.

Related Skills

Explore More Software Engineering Skills→
curtismu7curtismu7
1

caveman

Compresses AI responses into terse caveman-style phrasing while preserving technical accuracy.

Community
Basic
Haohao-endHaohao-end
779

caveman

Compress responses to retain conclusions, constraints, next steps, and file references.

Community
Basic
SaltFishGCSaltFishGC

caveman

Compress assistant responses into brief caveman-style prose with adjustable brevity levels.

Community
Intermediate
47network47network
5

prompt-compressor

Compress long prompts and contexts while preserving semantic meaning and reducing token usage.

Community
Intermediate