token-optimizerlisted
Install: claude install-skill samibajwaisking/Token-Optimizer-Skill
# Token Optimizer — Maximum Output, Minimum Input
A skill for extracting the highest-quality, most complete responses from Claude using
the fewest possible input tokens. Every technique here is tested, practical, and immediately
applicable.
---
## Core Philosophy
> **Information Density > Verbosity**
> Claude processes meaning, not words. A 40-token instruction that is precise always
> beats a 200-token instruction that is vague.
The goal is not to "trick" Claude — it is to communicate with maximum clarity and
minimum redundancy so Claude spends its compute on output, not parsing.
---
## The 5 Pillars of Token Optimization
### Pillar 1 — Prompt Compression
Strip all filler. Every word must earn its place.
**Before (47 tokens):**
```
Could you please analyze the following code and let me know what might be wrong
with it and also suggest how I could fix the issues you find?
```
**After (9 tokens):**
```
debug+fix: [code]
```
**Compression patterns:**
| Verbose Phrase | Compressed Form |
|---|---|
| "Please analyze and explain" | `analyze:` |
| "Write a detailed explanation of" | `explain:` |
| "Can you help me understand" | `clarify:` |
| "Please review and improve" | `refine:` |
| "Give me a step by step guide" | `steps:` |
| "Summarize the key points of" | `summarize:` |
| "Compare X and Y in detail" | `compare: X vs Y` |
| "Generate multiple options for" | `options(5):` |
Read `references/compression-patterns.md` for 50+ compression shortcuts across categories.