A professional-grade editorial system for Chinese content that automates proofreading, fact-checking, and regulatory compliance through an 8-layer pipeline.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install editorial-review
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 editorial-review 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 Editorial Review?
The Editorial Review skill is a sophisticated quality assurance framework designed for any public-facing Chinese text. Modeled after the prestigious "Three Reviews, Three Proofs" publishing industry standard, it provides a comprehensive single-pass analysis to eliminate embarrassing typos, grammatical errors, and factual inaccuracies. Beyond simple spell-checking, it serves as a critical safety net for organizations by identifying high-stakes risks like political sensitivity and advertising law violations.
By integrating this tool into your workflow, you ensure that every piece of content—from social media posts to official press releases—maintains a high standard of professional integrity. As a standout entry in the Openclaw Skills ecosystem, it delivers structured reports that categorize issues by severity, providing actionable fixes that protect your brand's reputation and ensure legal compliance in the Chinese market.
Editorial Review Use Cases
- Pre-Publication Compliance: Verifying that marketing copy adheres to strict Chinese advertising laws regarding superlatives and absolute claims.
- Political Sensitivity Screening: Ensuring territorial and political references are accurate and compliant with regional standards.
- Academic and News Proofreading: Detecting homophone errors and shape-similar character mistakes that standard spell-checkers miss.
- Technical Documentation Review: Maintaining consistent terminology and proper CJK-Latin typography (spacing between Chinese and English characters).
How Editorial Review Works
- Content Ingestion: The agent analyzes the submitted Chinese text to determine its length, target audience, and publication context.
- Mode Selection: Based on content stakes, the agent selects between Quick, Standard, or Deep review modes to apply the appropriate depth of the Openclaw Skills pipeline.
- Sequential 8-Layer Processing: The text undergoes a rigorous check through layers ranging from Character Accuracy and Punctuation to Political Compliance and Readability.
- Issue Documentation: Every identified error is logged with its exact location, original text, suggested fix, and a severity rating (Critical, Major, or Minor).
- Report Generation: The system produces a structured Editorial Review Report (审校报告) summarizing findings and providing an overall readiness assessment.
Editorial Review Setup
To utilize this skill within your AI environment, ensure your agent has access to the core markdown and reference files. For CLI-based agents like Claude Code, you can initiate a review with a simple command:
# Example of triggering a review via CLI
openclaw run editorial-review --file="article.md" --mode="standard"
Ensure that the references/ directory containing typography, grammar, and compliance rules is present in your project root.
Editorial Review Data Schema & Taxonomy
The skill organizes its output into a structured report to ensure clarity and professional hand-off. The taxonomy is defined as follows:
| Component | Description |
|---|---|
| Overview | Includes content type, word count, and review mode. |
| Summary | Aggregated counts of Critical, Major, and Minor issues. |
| Layer Tables | Detailed breakdowns for each of the 8 layers containing locations and fixes. |
| Severity | A three-tier classification: Critical (must fix), Major (should fix), Minor (optional). |
| Clean Version | An optional full-text output with all suggested edits applied. |
Editorial Review Advanced Features
- Multi-Mode Scaling: Automatically escalates to Deep Review for content over 5,000 characters or high-stakes regulatory filings.
- Platform Mapping: Native compatibility with various tools including
AskUserQuestionfor Claude Code andrequest_user_inputfor Codex. - Regulatory Guardrails: Built-in prohibited word lists for Chinese Advertising Law to prevent massive financial fines.
- Context-Aware Typography: Specifically handles the nuances of GB/T 15834-2011 punctuation standards and full-width/half-width character mixing.
SKILL.md
Loading
Related Openclaw Skills
Code Review Assistant
An automated intelligent agent that analyzes source code to provide detailed review comments, performance tips, and security vulnerability detection.
QLD Property Research
An automated research tool for conducting comprehensive Queensland property due diligence using verified official government and council data sources.
Deep Research
An advanced research pipeline that decomposes complex queries into multi-round searches to generate structured reports with verified citations.
Jetson CUDA
A specialized toolkit for managing and querying NVIDIA CUDA environments and GPU device information on Jetson platforms.
ClawDNA -- Identity Backup & Sync
ClawDNA is a specialized version control and synchronization engine designed to protect and manage the core identity files of AI agents.
tianshu-xhs-note
A specialized tool for transforming raw text into structured Little Red Book (XHS) posts with optimized headings, layouts, and hashtags.
Featured*
