with one click
with one click
三层内容验证系统,确保内容质量符合平台要求
5维度质量评分和爆款标题生成,评估内容整体质量
Generate professional technical diagrams (Mermaid, ASCII, flowcharts, architecture diagrams) for AI/ML articles. Supports neural network architectures, data pipelines, algorithm flows, and system architectures.
When the user wants to edit, review, or improve existing marketing copy. Also use when the user mentions 'edit this copy,' 'review my copy,' 'copy feedback,' 'proofread,' 'polish this,' 'make this better,' or 'copy sweep.' This skill provides a systematic approach to editing marketing copy through multiple focused passes.
When the user wants to write, rewrite, or improve marketing copy for any page — including homepage, landing pages, pricing pages, feature pages, about pages, or product pages. Also use when the user says "write copy for," "improve this copy," "rewrite this page," "marketing copy," "headline help," or "CTA copy." For email copy, see email-sequence. For popup copy, see popup-cro.
When the user wants to create or optimize an email sequence, drip campaign, automated email flow, or lifecycle email program. Also use when the user mentions "email sequence," "drip campaign," "nurture sequence," "onboarding emails," "welcome sequence," "re-engagement emails," "email automation," or "lifecycle emails." For in-app onboarding, see onboarding-cro.
| name | de-ai-humanizer |
| description | 四层去AI化处理,让内容更自然、有人情味,降低AI痕迹评分 |
| triggers | ["去AI化","人性化","去AI腔","内容自然化","AI痕迹","AI味"] |
将AI生成的内容转化为更自然、更有人情味的表达,通过四层处理降低"AI味"评分。
from src.utils.deai import DeAIAgent, humanize_content, detect_ai_traces
# 方式1: 使用类(完整控制)
agent = DeAIAgent()
humanized, report = agent.humanize(
content,
intensity=0.5, # 处理强度 0.0-1.0
platform="weixin", # 目标平台
enable_colloquial=True, # 启用口语化
enable_emotional=True, # 启用情感注入
enable_particles=True # 启用语气词
)
print(f"AI味评分: {report.original_ai_score}% → {report.final_ai_score}%")
print(f"连接词替换: {len(report.connector_replacements)} 处")
print(f"口语化替换: {len(report.colloquial_replacements)} 处")
# 方式2: 使用便捷函数
humanized, report = humanize_content(content, platform="xiaohongshu")
# 方式3: 检测AI痕迹
traces = detect_ai_traces(content)
for trace in traces:
print(f"发现: {trace['text']} (出现 {trace['count']} 次)")
将AI痕迹明显的连接词替换为自然表达:
| AI连接词 | 人性化替代词 |
|---|---|
| 首先 | 说起来、其实吧、老实说、一开始 |
| 其次 | 还有一点、另外啊、对了、而且 |
| 综上所述 | 说到底、归根结底、简单说 |
| 然而 | 不过、但是、可、话说回来 |
| 值得注意的是 | 有意思的是、你知道吗、重点来了 |
将正式词汇转为口语化表达:
| 正式词 | 口语化替代 |
|---|---|
| 认为 | 觉得、感觉、寻思、想 |
| 使用 | 用、整、搞、弄 |
| 实现 | 搞定、弄好、做到 |
| 非常 | 特别、超、老、贼 |
| 进行 | 做、搞、弄、整 |
添加个人感受和情绪词:
# 情感词类型
EMOTION_WORDS = {
"positive_surprise": ["真的没想到", "最让我惊喜的是", "太赞了"],
"question": ["你有多久没...了?", "你猜怎么着?"],
"exclamation": ["这个真的太...了!", "关键是"],
"story": ["我之前也是", "后来我发现", "让我印象深的是"],
"empathy": ["这真不是你一个人的问题", "我也一样"],
}
添加"吧、呢、啊"等语气词:
| 语气词 | 情感 | 用法 |
|---|---|---|
| 啊 | 感叹 | 强调、惊讶 |
| 呢 | 亲切 | 轻松、延续 |
| 吧 | 缓和 | 推测、建议 |
| 哦 | 提醒 | 恍然大悟 |
| 嘛 | 解释 | 明显、解释 |
不同平台有不同的去AI化强度:
| 平台 | intensity | colloquial | emotional | 目标AI味 |
|---|---|---|---|---|
| 微信公众号 | 0.7 | ✅ | ✅ | ≤15% |
| 小红书 | 0.3 | ✅ | ❌ | ≤30% |
| Twitter/X | 0.2 | ❌ | ❌ | ≤40% |
| 知乎 | 0.4 | ✅ | ❌ | ≤20% |
# 微信公众号:高强度去AI化
humanized, report = agent.humanize(content, platform="weixin", intensity=0.7)
# 小红书:适度口语化,不添加情感
humanized, report = agent.humanize(content, platform="xiaohongshu", intensity=0.3)
# Twitter:保持专业,轻度处理
humanized, report = agent.humanize(content, platform="twitter", intensity=0.2)
# 评分公式
ai_score = (AI连接词字数 * 2 + 禁忌标题字数 * 4) / 总字数 * 100
# 评分标准
# ≤15% 优秀 - 内容自然,AI痕迹很少
# 15-30% 良好 - 轻微AI痕迹,可接受
# >30% 需改进 - AI痕迹明显,建议处理
检查并建议小标题格式:
# 禁忌格式
forbidden = ["第一章", "1.", "一、", "第一部分", "第一节"]
# 推荐格式
recommended = {
"question": "提问式:为什么需要关注XXX?",
"number": "数字式:3个关键步骤掌握XXX",
"contrast": "对比式:传统方法 vs 创新方案",
"golden": "金句式:破解XXX的核心秘密",
"benefit": "利益式:如何用XXX提升效率",
"painpoint": "痛点式:别再为XXX烦恼了",
}
# 使用验证
issues = agent.validate_subheadings(content)
suggestions = agent.suggest_subheadings("AI写作", count=5)
from src.utils.deai import DeAIAgent
# 原始内容(AI痕迹明显)
content = """
首先,我们需要了解AI的基本原理。其次,AI的应用非常广泛。
综上所述,AI是一个值得深入研究的领域。
"""
# 创建处理器
agent = DeAIAgent()
# 检测AI痕迹
traces = agent.detect_ai_patterns(content)
# 输出: [
# {"type": "ai_connector", "text": "首先", "count": 1},
# {"type": "ai_connector", "text": "其次", "count": 1},
# {"type": "ai_connector", "text": "综上所述", "count": 1}
# ]
# 去AI化处理
humanized, report = agent.humanize(content, platform="weixin")
print(f"原始AI味: {report.original_ai_score}%")
print(f"处理后AI味: {report.final_ai_score}%")
print(f"替换详情: {report.connector_replacements}")
# 输出示例:
# 原始AI味: 8.5%
# 处理后AI味: 2.1%
# 替换详情: [
# {"original": "首先", "replacement": "说起来"},
# {"original": "其次", "replacement": "还有一点"},
# {"original": "综上所述", "replacement": "说到底"}
# ]
Created: 2026-02-14
Version: 1.0
Dependencies: src/utils/deai/