A sophisticated text compressor that converts modern Chinese or English into concise Classical Chinese to significantly reduce LLM token usage.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install wenyan-compress
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install wenyan-compress using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is wenyan-compress?
wenyan-compress is a specialized tool designed for the Openclaw Skills ecosystem to tackle the challenge of high token consumption in large language models. By leveraging the inherent brevity of Classical Chinese (Wenyanwen), it transforms verbose modern prose and English text into compact, information-dense summaries. This process not only saves costs but also allows more information to fit within a model's context window.
The skill provides multiple compression styles—ranging from poetic to ultra-short—and includes a robust fidelity check mechanism. This ensures that while the text is shortened, the core meaning, critical numbers, and proper nouns remain intact, making it an essential utility for developers and power users managing long-form content.
wenyan-compress Use Cases
- Reducing token counts for long-form articles or documentation to lower API costs.
- Summarizing meeting minutes or product requirement documents (PRD) into concise bullet points.
- Creating stylistic, ancient-style content or literary summaries for creative writing.
- Extracting core insights from academic papers or news reports while preserving key data.
- Managing context window limits for AI agents by condensing previous conversation history.
How wenyan-compress Works
- The user inputs text (English or Chinese) and optional command flags to the skill.
- For English input, the system performs a two-step transformation: translating to Modern Chinese first, then to Classical Chinese to ensure accuracy.
- The skill applies the selected compression logic (Extreme, Normal, Detailed, or Poetic) based on the user's priority.
- A fidelity check identifies key entities like dates, names, and numbers to verify they are preserved in the output.
- The skill generates a final compressed output accompanied by a fidelity report and token savings statistics.
wenyan-compress Setup
To use this skill within the Openclaw Skills environment, ensure you have the latest version installed. You can trigger the compression by passing your text directly to the agent:
# Example of using the skill with a fidelity flag
openclaw run wenyan-compress "Your long text here... !fidelity"
wenyan-compress Data Schema & Taxonomy
The skill organizes its output and metadata as follows:
| Data Point | Description | Example |
|---|---|---|
| Output Text | The final Classical Chinese version of the input. | AI变世道，机器学于历... |
| Fidelity Report | Validation of retained key information and risk alerts. | ✅ Retained: Dates, Names |
| Compression Ratio | Comparison between original and compressed token counts. | 5.4 : 1 |
| Savings Percentage | Estimated percentage of tokens saved. | ~81% |
Available command flags:
!detailed: Higher retention (~95% detail).!extreme: Maximum compression (~70% retention).!poetic: Stylized literary output.!fidelity: High-priority mode for preserving technical terms and numbers.
wenyan-compress Advanced Features
- Automatic Language Detection: Seamlessly handles both English and Chinese inputs without manual switching.
- Multi-Level Fidelity System: Users can prioritize precision for technical data or brevity for general summaries.
- Fidelity Verification Hook: Automatically scans compressed output for potential information loss and warns the user.
- Token Analytics: Real-time calculation of token savings to help users monitor efficiency within Openclaw Skills workflows.
SKILL.md
Loading
Related Openclaw Skills
todo-sync
An automated synchronization bridge that keeps tasks consistent between Enterprise WeChat and Feishu platforms.
WAF Rule Validator
A comprehensive security assessment tool designed to test the effectiveness of Web Application Firewalls and API gateways through malicious request simulation.
Auto File Organizer
An intelligent AI agent skill designed to automatically categorize and tidy cluttered directories like download folders.
Court Notice Automation
An automated workflow for parsing court notices, extracting key trial data, and synchronizing it with macOS calendars and reminders.
Novel Assistant
A specialized AI writing tool focused on maintaining narrative consistency, tracking complex plot threads, and managing multi-novel projects.
Math Review Generator
An automated engine for creating comprehensive math review materials in HTML format featuring validated SVG visualizations and original practice questions.
Featured*
