Humanize CLI

用于检测并自动修正AI写作模式的命令行工具。

脚本

analyze.py — 检测AI模式

扫描文本并报告AI常用词汇、夸大表述、聊天机器人痕迹以及可自动替换的短语。

# Analyze a file
python scripts/analyze.py input.txt

# Analyze from stdin
echo "This serves as a testament to our commitment" | python scripts/analyze.py

# JSON output for programmatic use
python scripts/analyze.py input.txt --json

输出示例：

==================================================
AI PATTERN ANALYSIS - 5 issues found
==================================================

AI VOCABULARY:
  • testament: 1x
  • crucial: 2x

AUTO-REPLACEABLE:
  • "serves as" → "is": 1x
  • "in order to" → "to": 1x

humanize.py — 自动替换模式

对常见的AI化表达执行自动替换。

# Humanize and print to stdout
python scripts/humanize.py input.txt

# Write to output file
python scripts/humanize.py input.txt -o output.txt

# Include em dash replacement
python scripts/humanize.py input.txt --fix-dashes

# Quiet mode (no change log)
python scripts/humanize.py input.txt -q

自动修正的内容：

• 填充短语："in order to" → "to", "due to the fact that" → "because"
• 系动词回避："serves as" → "is", "boasts" → "has"
• 句子开头词：移除 "Additionally,", "Furthermore,", "Moreover,"
• 弯引号 → 直引号
• 聊天机器人痕迹：移除 "I hope this helps", "Let me know if" 等

工作流程

1. 先进行分析以查看需要修正的内容：

   python scripts/analyze.py document.txt
   

2. 自动修正安全的替换：

   python scripts/humanize.py document.txt -o document_clean.txt
   

3. 手动审核针对分析标记的AI术语和夸大表述（这些需要人工判断）

4. 重新分析以确认改进效果：

   python scripts/analyze.py document_clean.txt
   

自定义模式

编辑scripts/patterns.json以添加/移除：

• ai_words— 仅标记但不自动替换的术语
• puffery— 需要标记的推广性语言
• replacements— 词组→替换映射（空字符串表示删除）
• chatbot_artifacts— 需自动移除的短语
• hedging_phrases— 需标记的过度模糊表述

批量处理

处理多个文件：

# Analyze all markdown files
for f in *.md; do
  echo "=== $f ===" 
  python scripts/analyze.py "$f"
done

# Humanize all txt files in place
for f in *.txt; do
  python scripts/humanize.py "$f" -o "$f.tmp" && mv "$f.tmp" "$f"
done