megacave skill: make Claude answer in Classical Chinese

Makes Claude answer in terse Classical Chinese (文言文), keeping code, commands and numbers exact.

Installs: 32K; on skills.sh
Stars: 116.7K; on GitHub
Files: 1; besides SKILL.md
Updated: Nov 2026; last repo commit

megacave is the Classical Chinese voice in the caveman family. Claude answers in 文言文: subjects dropped where clear, verbs before objects, and classical particles instead of connecting phrases. The reply on screen gets far shorter. Technical terms, code, commands, paths, error text and numbers stay in their original script.

It starts only when you type /megacave or /caveman wenyan, and stays on until you say "stop caveman" or "normal mode". It switches to plain modern Chinese or your own language for security warnings, irreversible actions, ambiguous clauses and anything saved outside the chat.

Good for

Readers who enjoy or work in Classical Chinese
Developers who want compact replies and can read 文言文 comfortably
People curious how far an assistant can compress a technical answer

Install megacave

Claude Code

Copy this prompt.

Install the megacave skill in Claude Code. Run this command:

npx skills add https://github.com/juliusbrussee/caveman --skill megacave

Then check that it loaded and tell me how to use it.

2
Paste it into Claude Code. It shows the change and asks before running it.

Requires Node.js.

From its listing on skills.sh. Source.

What's in the skill

When it loads

Caveman in Classical Chinese: 文言文 register, far fewer characters, technical terms verbatim. Invoke only with /megacave or /caveman wenyan. Stays on until "stop caveman" or "normal mode".

Sections in SKILL.md

Persistence
Floor
Rules
1. Classical register
2. Particles carry structure
3. Each fact once
4. Payload in original script
5. Tool runs
6. Never perform
When to break the rules
Pre-send check

Key facts

Author: juliusbrussee
Repo: juliusbrussee/caveman
Installs on skills.sh: 32,033 as of Nov 2, 2026
GitHub stars: 116,702 for the whole juliusbrussee/caveman repo, Nov 2, 2026
License: Apache-2.0 (crawl data; the repo README also says Apache-2.0, whole repo)
Last repo commit: Nov 1, 2026
What is inside: 2 files: SKILL.md and README.md
skills.sh audits: Gen Agent Trust Hub pass, Socket pass, Snyk pass
How it starts: Only when you type /megacave (or its /caveman alias); the SKILL.md says to invoke it only that way
Stays on: Until you say "stop caveman" or "normal mode"
Part of: The caveman repo, next to the base caveman skill

What it includes

Turns on the Classical Chinese voice for the whole session
Also reachable as /caveman wenyan, which the README calls an alias.
Try: “/megacave”
/megacave
Reports the current mode and changes nothing
/caveman status
文言, not 白話: verb before object, subjects omitted where recoverable
Rule 1.
Classical register
One particle, one relation (之, 乃, 為, 其, 則, 故, 以)
Rule 2.
Particles carry structure
No restating; for code changes show changed lines plus context
Rule 3.
Each fact once
Identifiers such as useMemo stay in Latin letters
Rule 4.
Payload in original script
Says nothing between tool calls unless something changes
Rule 5.
Tool runs: one line in, one line out
Code, commands, paths, API names, errors, negations and numbers stay as they are
Numbers use Arabic numerals. A clause that is ambiguous in 文言 becomes 白話.
The floor
Security warnings, irreversible actions, confusion, questions to you
Also anything persisted outside chat.
When it speaks plainly
Say "stop caveman" or "normal mode" to end it
Stop

From the official docs

Questions

What does megacave do?

It makes Claude answer in Classical Chinese (文言文), with far fewer characters on screen. Code, commands, paths, error messages and numbers stay exactly as they are.

How do I install megacave?

Use the install box at the top of this page. Then type /megacave to start. The /caveman wenyan shortcut comes from the base caveman skill, so it works only if that is installed too.

Do I need to read Classical Chinese?

Yes, to follow the replies. When a clause would be ambiguous in 文言文 the skill switches to plain modern Chinese for that clause. Say "stop caveman" to go back to normal answers.

Does megacave save tokens?

Only a little. The SKILL.md itself says token savings are small and noisy and must not be claimed. The README measured 9% fewer output tokens at the median against a plain "Answer concisely." instruction, ranging from 12% more to 39% fewer, on ten prompts and a single run.

How do I turn megacave off?

Say "stop caveman" or "normal mode". The skill stays on for every reply until you do.

megacave vs ultracave vs caveman?

All three shorten answers. Caveman and ultracave stay in English or your own language. Megacave switches to Classical Chinese, so the text is short on screen but not necessarily in tokens.

Is megacave safe?

The skill is two text files, SKILL.md and README.md, and bundles no scripts. The main caveman README says the skill runs on your machine and sends nothing, and that its separate command line installer sends usage stats unless you turn them off with caveman telemetry off or DO_NOT_TRACK=1.

Information

updated Nov 2, 2026

Repository: juliusbrussee/caveman
Folder: skills/megacave

juliusbrussee/caveman skills.sh

Sources

Docs, issue trackers and discussions this page is built from.

Show all 3

megacave SKILL.md · github.com
megacave README · github.com
skills.sh listing · skills.sh