Rules

What Is This

The "Rules" skill for the Happycapy Skills platform (Skill ID: caveman) is an ultra-compressed communication mode designed to reduce token usage by approximately 75 percent. It achieves this by formatting responses in a style reminiscent of "caveman speak" while maintaining full technical accuracy. The skill processes language to remove unnecessary words, simplify phrasing, and condense information without sacrificing clarity or correctness in technical contexts. It supports three intensity levels-lite, full (default), and ultra-which determine how aggressively the language is compressed.

The skill is activated by specific user prompts such as "caveman mode", "talk like caveman", "be brief", "less tokens", or the /caveman command. It can also auto-trigger when users request token efficiency. The source code and configuration details can be found at the caveman GitHub repository.

Why Use It

Token usage is a central concern in environments where communication is mediated by language models with strict token limits or cost constraints. Excessive verbosity, filler words, and formalities can quickly inflate token counts, leading to truncated responses, higher costs, or slower processing. The "Rules" skill directly addresses this by stripping responses down to their technical essence, allowing for:

• Maximized information density per token.
• Reduced communication costs.
• Faster exchanges, especially in chat or API contexts.
• Minimalist, distraction-free technical discussions.

This mode is especially useful for developers, technical teams, and anyone working within environments where bandwidth, cost, or token efficiency is a priority.

How to Use It

Activating the "Rules" skill is straightforward. Users can invoke it in several ways:

• Explicit command: /caveman
• Specifying intensity: /caveman lite, /caveman full, /caveman ultra
• Natural language triggers: "caveman mode", "talk like caveman", "use caveman", "less tokens", "be brief"
• Automatic activation: When the system detects a request for token efficiency

Intensity Levels

Switching Intensity Example:

/caveman ultra

Rules of Compression

• Drop:

  • Articles (a, an, the)
  • Fillers (just, really, basically, actually, simply)
  • Pleasantries (sure, certainly, of course, happy to)
  • Hedging language

• Keep:

  • Full technical substance
  • Exact technical terms
  • Code blocks unchanged
  • Exact quoting of errors

• Rewrite:

  • Use fragments and short synonyms (e.g., "fix" instead of "implement a solution for")
  • Use the pattern: [thing] [action] [reason]. [next step].
  • For ultra mode: abbreviate terms, remove conjunctions, use arrows for cause-effect, use single words where possible

Example Transformation:

• Not: "Sure! I'd be happy to help you with that. The issue you're experiencing is likely caused by..."
• Yes: "Bug in auth middleware. Token expiry check use < not <=. Fix:"

Code Example: Input:

Can you show me how to handle 404 errors in Express?

Output (full):

Express: add middleware after routes. If req not matched, send 404.

```js
app.use((req, res) => {
  res.status(404).send('Not Found');
});

Output (ultra):
```text
Express: 404 → last middleware.

```js
app.use((req, res) => res.status(404).send('Not Found'));

## When to Use It

- When token efficiency is critical (e.g., working with limited-token LLMs)
- In high-frequency chat or API scenarios where brevity is required
- When conveying technical concepts without the need for formal or verbose language
- During code reviews, debugging, or technical troubleshooting where density is prioritized
- When requested by the user for "caveman mode" or similar triggers

## Important Notes

- The "Rules" skill never removes technical accuracy or alters code blocks-only linguistic compression occurs.
- Use caution in environments where ultra-abbreviated language may hinder team communication or documentation clarity.
- The default mode is "full", which strikes a balance between brevity and readability. Use "ultra" only where maximum compression is required.
- Responses will always prioritize technical substance, ensuring that no critical information is lost in the compression process.
- Skill is English-only and does not support other languages or non-ASCII symbols.
- All outputs comply with Happycapy’s guidelines for professional, technically accurate communication and strict English-only output. No Chinese characters, no emojis, no em-dashes.

The "Rules" skill is ideal for technical users who demand high efficiency and uncompromising accuracy in communication, especially where token economy is a primary concern.