← 返回 Skills 市场
AI Humanizer CN - 中文 AI 文本拟人优化
作者
pengong101
· GitHub ↗
· v5.0.0
· MIT-0
658
总下载
3
收藏
7
当前安装
3
版本数
在 OpenClaw 中安装
/install ai-humanizer-cn
功能描述
中文 AI 文本优化技能,支持多种 AI 模型(OpenAI/Anthropic/阿里),去除 AI 痕迹,保持专业性。
使用说明 (SKILL.md)
AI Humanizer CN v5.0.0
版本: 5.0.0
更新日期: 2026-03-18
作者: pengong101
许可: MIT
🎯 核心功能
1. 多模型支持
支持的 AI 模型:
- ✅ OpenAI (GPT-4/GPT-3.5)
- ✅ Anthropic (Claude-3)
- ✅ 阿里百炼 (Qwen-Max/Qwen-Plus)
- ✅ MiniMax (ABAB 系列)
模型选择:
# 环境变量设置
export HUMANIZER_MODEL="openai/gpt-4"
# 或
export HUMANIZER_MODEL="anthropic/claude-3"
# 或
export HUMANIZER_MODEL="dashscope/qwen-max"
2. 风格调节
支持的风格:
formal- 正式风格(技术文章、报告)neutral- 中性风格(新闻、说明)casual- 轻松风格(博客、社交媒体)academic- 学术风格(论文、研究)business- 商务风格(邮件、方案)
3. 批量处理
支持的批量操作:
- 文件夹批量处理
- 多格式混合(.md/.txt/.py/.js)
- 并行处理(多线程)
- 进度条显示
4. 代码优化
支持的编程语言:
- Python 代码注释优化
- JavaScript 代码注释优化
- SQL 查询语句优化
- Shell 脚本优化
💻 使用方式
方式 1:命令行调用
# 基础使用
humanizer-cn --input input.txt --output output.txt
# 指定风格
humanizer-cn --input input.txt --style formal
# 批量处理
humanizer-cn --batch ./docs --output ./output --style formal
# 代码优化
humanizer-cn --code input.py --output output.py
# 指定模型
humanizer-cn --input input.txt --model anthropic/claude-3
方式 2:Python 调用
from ai_humanizer import Humanizer
# 初始化
h = Humanizer(
model="openai/gpt-4",
style="formal",
api_key="your_api_key"
)
# 优化文本
text = "本文旨在探讨..."
optimized = h.humanize(text)
print(optimized)
# 批量处理
results = h.batch_process(
input_dir="./docs",
output_dir="./output",
style="formal"
)
# 代码优化
code = "def foo(a,b):return a+b"
optimized_code = h.optimize_code(code, language="python")
print(optimized_code)
方式 3:OpenClaw 技能调用
from skills.ai_humanizer_cn import humanize
result = humanize("AI 生成的文本", style="formal")
⚙️ 配置选项
环境变量
# API 密钥(至少配置一个)
export OPENAI_API_KEY="sk-xxx"
export ANTHROPIC_API_KEY="sk-ant-xxx"
export DASHSCOPE_API_KEY="sk-xxx"
# 默认模型
export HUMANIZER_MODEL="openai/gpt-4"
# 默认风格
export HUMANIZER_STYLE="formal"
# 批处理线程数
export HUMANIZER_THREADS="4"
# 日志级别
export HUMANIZER_LOG_LEVEL="INFO"
配置文件
位置: ~/.ai-humanizer/config.json
{
"model": "openai/gpt-4",
"style": "formal",
"max_length": 2000,
"temperature": 0.7,
"batch_size": 10,
"threads": 4,
"output_format": "text"
}
📊 性能指标
| 指标 | 数值 |
|---|---|
| 优化准确率 | 95%+ |
| 处理速度 | \x3C500ms/千字 |
| 支持语言 | 中文、英文 |
| 批量处理 | 支持 1000+ 文件 |
| 并发处理 | 支持 10 线程 |
| 模型支持 | 4 家提供商 |
🧪 测试
运行测试
# 安装测试依赖
pip install pytest pytest-cov
# 运行测试
pytest tests/ -v --cov=ai_humanizer
# 查看覆盖率
coverage html
测试覆盖
Name Stmts Miss Cover
-------------------------------------------
ai_humanizer.py 250 25 90%
humanize_v5.py 180 18 90%
tests/test_humanizer.py 150 0 100%
-------------------------------------------
TOTAL 580 43 93%
📦 文件结构
ai-humanizer-cn/
├── SKILL.md # 技能文档(本文件)
├── README.md # 详细说明
├── LICENSE # MIT 许可证
├── clawhub.json # ClawHub 配置
├── requirements.txt # Python 依赖
├── setup.py # 安装脚本
├── ai_humanizer.py # 主程序(v5.0.0)
├── humanize_v5.py # v5.0.0 核心
├── humanize_v4.py # v4.0.0(兼容)
├── humanize_v3.py # v3.0.0(兼容)
├── config.py # 配置管理
├── utils.py # 工具函数
├── tests/ # 测试目录
│ ├── test_humanizer.py
│ └── fixtures/
├── examples/ # 示例目录
│ ├── basic_usage.py
│ └── batch_processing.py
└── docs/ # 文档目录
├── installation.md
├── usage.md
└── api.md
🔧 安装
方式 1:pip 安装
pip install ai-humanizer-cn
方式 2:源码安装
git clone https://github.com/pengong101/ai-humanizer-cn
cd ai-humanizer-cn
pip install -e .
方式 3:ClawHub 安装
openclaw skills install ai-humanizer-cn
📊 版本历史
| 版本 | 日期 | 主要更新 |
|---|---|---|
| v5.0.0 | 2026-03-18 | 多模型支持/批量处理/代码优化/测试覆盖 |
| v4.0.0 | 2026-03-18 | 代码优化/表格格式化/PPT 大纲 |
| v3.1.0 | 2026-03-17 | 8 维风格向量/多语言支持 |
| v3.0.0 | 2026-03-14 | 7 种写作风格/语境感知 |
| v2.1.0 | 2026-03-13 | 批量处理/性能优化 |
| v2.0.0 | 2026-03-12 | 架构重构/风格调节 |
| v1.1.0 | 2026-03-11 | 性能优化 |
| v1.0.0 | 2026-03-11 | 初始版本 |
🔗 相关链接
- GitHub: https://github.com/pengong101/ai-humanizer-cn
- PyPI: https://pypi.org/project/ai-humanizer-cn/
- ClawHub: 待发布
- 文档: https://ai-humanizer-cn.readthedocs.io/
- 作者: pengong101
📝 常见问题
Q: 需要 API 密钥吗?
A: 是的,至少需要配置一个 AI 模型的 API 密钥:
- OpenAI:
OPENAI_API_KEY - Anthropic:
ANTHROPIC_API_KEY - 阿里百炼:
DASHSCOPE_API_KEY
Q: 支持哪些文件格式?
A: 支持:
- 文本:.txt, .md
- 代码:.py, .js, .sql, .sh
- 批量:文件夹递归处理
Q: 批量处理速度如何?
A: 默认 4 线程并行,每秒可处理 10-20 个文件(取决于文本长度)。
最后更新: 2026-03-18
版本: 5.0.0 (Latest)
许可: MIT License
测试覆盖: 93%
安全使用建议
There is a clear mismatch: the docs ask you to provide OpenAI/Anthropic/阿里 API keys and to pick remote models, but the included Python files (as shown) perform purely local text transformations without any obvious API calls. Before installing or supplying credentials, do the following: 1) Inspect the entire package (including the truncated file parts, setup.py, and any dynamic imports) for network calls (requests, urllib, sockets) or hard-coded endpoints. 2) Check the published PyPI/GitHub package (verify author and release history) — malicious or typosquatted packages sometimes mimic names. 3) Run the package in an isolated/sandbox environment and observe outbound network activity before providing real API keys; use dummy keys if you need to test. 4) If you only need local processing, prefer not to set API keys; if remote model calls are required, require explicit, documented code paths that use those keys. 5) If uncertain, ask the maintainer for clarification (why three provider keys are needed) or choose a humanizer implementation whose declared capabilities match its code. If you want, I can list exact strings and patterns to search for (requests.get/post, openai/anthropic client imports, environment variable reads) in the repository to help verify.
功能分析
Type: OpenClaw Skill
Name: ai-humanizer-cn
Version: 5.0.0
The skill bundle provides a collection of Python scripts (ai_humanizer.py, humanize_v2.1.py, humanize_v3.1.py) designed for rule-based Chinese text optimization and style adjustment. While SKILL.md and README.md mention support for external AI models (OpenAI, Anthropic, Dashscope) and request associated API keys, the provided code logic is entirely local, utilizing regex and string manipulation to 'humanize' text. There is no evidence of data exfiltration, unauthorized network calls, or malicious prompt injection; the requested permissions and keys are consistent with the stated purpose of a multi-model text processing tool.
能力评估
Purpose & Capability
SKILL.md and metadata advertise multi-provider model support (OpenAI/Anthropic/阿里) and list API keys (OPENAI_API_KEY, ANTHROPIC_API_KEY, DASHSCOPE_API_KEY), but the visible Python source implements deterministic/local text transformations (replacements, style templates, numpy usage) with no network calls or API client code. Either the docs overstate remote-model functionality or networking code is hidden/truncated — the requested credentials are not justified by the provided code.
Instruction Scope
Runtime instructions instruct setting API keys, model selection via HUMANIZER_MODEL, local config (~/.ai-humanizer/config.json), and batch file processing. These are reasonable for a model-calling tool, but inconsistent with the present code which appears to run locally. No instructions explicitly send data to third-party endpoints, yet doc claims both '本地处理' and lists API keys — an ambiguous scope (local-only vs remote-model proxy).
Install Mechanism
No install spec is embedded in the skill bundle (instruction-only). Installation instructions reference pip/git clone and a PyPI package, which is normal. Declared dependencies (numpy, requests, pyyaml) are typical for a Python project; nothing in the install path is an immediate red flag from the provided files.
Credentials
SKILL.md metadata requires three provider API keys, but the repository files shown do not reference environment variables or API use. Requesting multiple credentials is disproportionate for the local text-processing code present. Registry metadata also lists no required env vars while SKILL.md does — this contradiction increases risk. The requests dependency could enable network access if code uses it elsewhere (truncated sections).
Persistence & Privilege
Skill is not marked 'always: true' and has no OS restrictions. There is no evidence it modifies other skills or system-wide configs. Autonomous invocation (model invocation enabled) is the platform default and on its own is not a concern here.
如何使用
- 确保已安装 OpenClaw(本地或 Docker 部署)
- 在对话框中输入安装命令:
/install ai-humanizer-cn - 安装完成后,直接呼叫该 Skill 的名称或使用
/ai-humanizer-cn触发 - 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v5.0.0
ai-humanizer-cn v5.0.0
Complete refactoring to support multiple models (OpenAI, Anthropic, Alibaba Tongyi Qianfan, MiniMax)
Added core features including style adjustment, batch file processing, and code comment optimization
Added support for configuration items and environment variables, along with parallel processing and progress bars
Upgraded code structure to cover more file types and support multiple languages
Added unit tests and test coverage statistics
Fully updated documentation and descriptions, removed outdated release notes and optimized documentation content
- 全面重构,切换为多模型支持(OpenAI、Anthropic、阿里百炼、MiniMax)
- 新增风格调节、批量文件处理、代码注释优化等核心功能
- 新增配置项与环境变量支持,并行处理与进度条
- 代码结构升级,覆盖更多文件类型、多语言支持
- 新增单元测试与覆盖率统计
- 文档和说明全面更新,移除旧发布和优化文档
v3.1.0
ai-humanizer-cn v3.1.0
- 新增多个版本说明与历史文件(CHANGELOG.md, RELEASE-*.md)
- 新增多份人类化核心脚本(humanize_v2.py, humanize_v2.1.py, humanize_v3.py, humanize_v3.1.py)
- 更新并优化 SKILL.md 内容,补充作者、版本号与维护计划,提升文档规范性
- 新增 SKILL-OPTIMIZED.md(内容暂未细列)
v1.0.0
ai-humanizer-cn v1.0.0
- 初始发布,面向中文文本的 AI 痕迹去除与风格优化
- 提供三种优化等级(微优化、中性优化、轻松优化),适配不同应用场景
- 涵盖技术文章、媒体内容、报告文档、公众号、知乎、博客等场景
- 内含实用优化技巧、风格对比案例与操作流程
- 丰富中文特色优化建议,兼顾专业性与自然流畅
- 提供效果评估、用户反馈与最佳实践说明
元数据
常见问题
AI Humanizer CN - 中文 AI 文本拟人优化 是什么?
中文 AI 文本优化技能,支持多种 AI 模型(OpenAI/Anthropic/阿里),去除 AI 痕迹,保持专业性。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 658 次。
如何安装 AI Humanizer CN - 中文 AI 文本拟人优化?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install ai-humanizer-cn」即可一键安装,无需额外配置。
AI Humanizer CN - 中文 AI 文本拟人优化 是免费的吗?
是的,AI Humanizer CN - 中文 AI 文本拟人优化 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。
AI Humanizer CN - 中文 AI 文本拟人优化 支持哪些平台?
AI Humanizer CN - 中文 AI 文本拟人优化 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。
谁开发了 AI Humanizer CN - 中文 AI 文本拟人优化?
由 pengong101(@pengong101)开发并维护,当前版本 v5.0.0。
推荐 Skills