anti-aigc

Rewrites Chinese formal documents to reduce AI-content detection scores.

Updated Aug 1, 2026
One-click install
npx skills add https://github.com/lov-b/Skills --skill anti-aigc-lov-b
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: anti-aigc
Source: https://github.com/lov-b/Skills/tree/main/skills/anti-aigc
Command: npx skills add https://github.com/lov-b/Skills --skill anti-aigc-lov-b

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Chinese formal documents written or assisted by AI often carry detectable patterns—classical abbreviations, rigid parallel structures, excessive dashes and colons—that trigger AIGC detection tools. This Skill systematically rewrites such documents to read like natural human writing while preserving factual content and formal tone. ## Core Features & Use Cases - Nine-dimension rewriting: Expands abbreviations (如需→如果需要), naturalizes pronouns (该→这个), adds missing 会 in conditional sentences, replaces dashes and colons, varies parallel structures, and balances formality. - Partial replacement strategy: Applies each transformation to only 60-80% of instances, preserving natural inconsistency instead of uniform machine-style output. - Built-in AIGC scoring: After rewriting, evaluates each paragraph's risk level (high/medium/low) with a 0-100 score, displays a per-section report table, and iterates on high-risk paragraphs until acceptable. - Use Case: You have a Chinese technical design document flagged as AI-generated. Run this Skill to rewrite it across all nine dimensions, then review the detection score report showing which sections still need attention. ## Quick Start Ask the assistant to perform anti-AIGC optimization on your Chinese document and show the detection score report.

Frequently Asked Questions about anti-aigc

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I reduce AI detection rates in Chinese documents?

Apply systematic rewriting across multiple dimensions: expand classical abbreviations like 如需 to 如果需要, replace 该 with 这个, add missing 会 in conditional sentences, and break up rigid parallel structures. Partial replacement of 60-80% of instances produces more natural results than full replacement.

What patterns make Chinese text look AI-generated?

Common AI markers include heavy use of classical abbreviations (即可, 无需, 均可), pronouns like 该 and 其, frequent em-dashes for explanations, highly symmetric parallel sentences, formulaic transitions like 值得注意的是 and 综上所述, and missing 会 in conditional clauses.

Does anti-AIGC rewriting change the document's content?

No. The rewriting only changes expression, not facts. It prohibits adding or removing paragraphs, altering document structure, or introducing new technical descriptions, ensuring the content remains accurate while the style becomes more human-like.

Why not replace all AI-style phrases at once?

Full replacement creates a new uniform pattern that is itself detectable and unnatural. Human writing mixes forms, so replacing only 60-80% of instances and keeping some original phrasing produces the natural inconsistency that detection tools associate with human authors.

Can this approach work for English documents?

No. The rewriting rules target Chinese-specific patterns such as classical abbreviations, 该/其 pronouns, and Chinese punctuation conventions. English documents would require a different set of linguistic transformation rules.