ilang-compress

Compress natural language prompts into I-Lang — AI-native structured instructions. 40-65% token savings.

3,880 stars

Best use case

ilang-compress is best used when you need a repeatable AI agent workflow instead of a one-off prompt.

Compress natural language prompts into I-Lang — AI-native structured instructions. 40-65% token savings.

Teams using ilang-compress should expect a more consistent output, faster repeated execution, less prompt rewriting.

When to use this skill

  • You want a reusable workflow that can be run more than once with consistent structure.

When not to use this skill

  • You only need a quick one-off answer and do not need a reusable workflow.
  • You cannot install or maintain the underlying files, dependencies, or repository context.

Installation

Claude Code / Cursor / Codex

$curl -o ~/.claude/skills/ilang-compress/SKILL.md --create-dirs "https://raw.githubusercontent.com/openclaw/skills/main/skills/adsorgcn/ilang-compress/SKILL.md"

Manual Installation

  1. Download SKILL.md from GitHub
  2. Place it in .claude/skills/ilang-compress/SKILL.md inside your project
  3. Restart your AI agent — it will auto-discover the skill

How ilang-compress Compares

Feature / Agentilang-compressStandard Approach
Platform SupportNot specifiedLimited / Varies
Context Awareness High Baseline
Installation ComplexityUnknownN/A

Frequently Asked Questions

What does this skill do?

Compress natural language prompts into I-Lang — AI-native structured instructions. 40-65% token savings.

Where can I find the source code?

You can find the source code on GitHub using the link provided at the top of the page.

Related Guides

SKILL.md Source

# I-Lang Compress

An AI-native prompt compression protocol created by a Chinese developer.

Compress natural language prompts into dense structured instructions that any AI understands natively. 40-65% token savings, zero training needed.

## Why I-Lang

Token is money. Every prompt you send to GPT/Claude/Gemini, you pay by token. I-Lang compresses your instructions into a fraction of the original size — AI reads it just as well, you pay less.

## How to compress

When the user asks to compress a prompt, convert it to I-Lang syntax following these rules.

### Syntax

Single operation: `[VERB:@ENTITY|mod1=val1,mod2=val2]`
Pipe chain: `[VERB1:@SRC]=>[VERB2]=>[VERB3:@DST]`
Each step receives previous output as @PREV.

### Available Verbs (62)

Data I/O: READ, WRIT, DEL, LIST, COPY, MOVE, STRM, CACH, SYNC, Π
Transform: Σ, Δ, φ, ∇, DEDU, ∂, CHNK, FLAT, NEST, λ, REDU, PIVT, TRNS, ENCD, DECD, ξ, ζ, EXPN, θ, FMT
Analysis: ψ, CLST, SCOR, BNCH, AUDT, VALD, CNT, μ, TRND, CORR, FRCS, ANOM
Generation: CREA, DRFT, PARA, EXTD, SHRT, STYL, TMPL, FILL
Output: Ω, DISP, EXPT, PRNT, LOG
Meta: VERS, HELP, DESC, INTR, SELF, ECHO, NOOP

### Modifiers (28)

tgt, src, dst, frm, to, scp, dep, rng, whr, mch, exc, lim, off, top, bot, fmt, lng, sty, ton, len, col, row, srt, grp, typ, enc, chr, cap

### Entities (14)

@R2, @COS, @GH, @DRIVE, @LOCAL, @WORKER, @CF, @SCREEN, @LOG, @NULL, @STDIN, @SRC, @DST, @PREV

### Compression Guidelines

- Output the compressed I-Lang instruction first, then a brief explanation of what each step does.
- Use pipe chains for multi-step operations.
- Use Greek symbols where applicable (Σ for merge, Δ for diff, φ for filter, etc.)
- Maximize compression while preserving complete semantics.
- If input is ambiguous, ask the user for clarification.

## Examples

**Input:** Read the config file from GitHub and format it as JSON
**Output:** `[READ:@GH|path=config.json]=>[FMT|fmt=json]`
**Explanation:** READ fetches from GitHub, FMT converts to JSON format.
**Saved:** 55%

**Input:** Filter all fatal errors from system logs
**Output:** `[φ:@LOG|whr="lvl=fatal"]`
**Explanation:** φ (filter) selects only entries matching fatal level.
**Saved:** 55%

**Input:** Read all markdown files, merge them, summarize in 3 bullets, output
**Output:** `[LIST:@LOCAL|mch="*.md"]=>[Π:READ]=>[Σ|len=3]=>[Ω]`
**Explanation:** LIST finds files, Π batch-reads, Σ summarizes to 3 items, Ω outputs.
**Saved:** 65%

## Links

- Homepage: https://ilang.ai
- Dictionary: https://github.com/ilang-ai/ilang-dict

## Author

Built by ilang-ai from China. I-Lang is open source under MIT license.

I-Lang v2.0

Related Skills

media-compress

3891
from openclaw/skills

Compress and convert images and videos using ffmpeg. Use when the user wants to reduce file size, change format, resize, or optimize media files. Handles common formats like JPG, PNG, WebP, MP4, MOV, WebM. Triggers on phrases like "compress image", "compress video", "reduce file size", "convert to webp/mp4", "resize image", "make image smaller", "batch compress", "optimize media".

General Utilities

per-agent-compression-universal

3891
from openclaw/skills

Zero-config memory consolidation for multi-agent OpenClaw deployments. Auto-discovers agents and registers staggered cron tasks with full state tracking, deduplication, and domain-aware extraction.

ehr-semantic-compressor

3891
from openclaw/skills

AI-powered EHR summarization using Transformer architecture to extract key clinical information from lengthy medical records

ppt-compress

3891
from openclaw/skills

Compress PPT/PPTX file size. Decompress PPT, compress large images, repackage and convert to PDF to significantly reduce file size. Suitable for scenarios where large PPT files need to be shared or uploaded.

bundled-compression-runtime

3795
from openclaw/skills

Compress or convert generated images for delivery, preview, and social upload. Use when the user asks to compress an image, optimize image size, convert to webp, or reduce file size after generation.

---

3891
from openclaw/skills

name: article-factory-wechat

Content & Documentation

humanizer

3891
from openclaw/skills

Remove signs of AI-generated writing from text. Use when editing or reviewing text to make it sound more natural and human-written. Based on Wikipedia's comprehensive "Signs of AI writing" guide. Detects and fixes patterns including: inflated symbolism, promotional language, superficial -ing analyses, vague attributions, em dash overuse, rule of three, AI vocabulary words, negative parallelisms, and excessive conjunctive phrases.

Content & Documentation

find-skills

3891
from openclaw/skills

Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.

General Utilities

tavily-search

3891
from openclaw/skills

Use Tavily API for real-time web search and content extraction. Use when: user needs real-time web search results, research, or current information from the web. Requires Tavily API key.

Data & Research

baidu-search

3891
from openclaw/skills

Search the web using Baidu AI Search Engine (BDSE). Use for live information, documentation, or research topics.

Data & Research

agent-autonomy-kit

3891
from openclaw/skills

Stop waiting for prompts. Keep working.

Workflow & Productivity

Meeting Prep

3891
from openclaw/skills

Never walk into a meeting unprepared again. Your agent researches all attendees before calendar events—pulling LinkedIn profiles, recent company news, mutual connections, and conversation starters. Generates a briefing doc with talking points, icebreakers, and context so you show up informed and confident. Triggered automatically before meetings or on-demand. Configure research depth, advance timing, and output format. Walking into meetings blind is amateur hour—missed connections, generic small talk, zero leverage. Use when setting up meeting intelligence, researching specific attendees, generating pre-meeting briefs, or automating your prep workflow.

Workflow & Productivity