← Back to Skills Marketplace
39
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install wechat-assistant
Description
微信聊天记录智能分析与自动回复助手;自动抓取聊天记录、智能分析挖掘需求、生成回复内容并支持审核发送;当用户需要分析微信聊天数据、自动生成回复、管理人脉关系或规划任务提醒时使用
README (SKILL.md)
微信聊天助手
任务目标
- 本Skill用于:Windows平台微信聊天记录的智能抓取、分析与自动回复
- 能力包含:聊天记录抓取、需求分析挖掘、知识库管理、回复生成与审核、数据可视化、任务规划
- 触发条件:用户需要分析微信聊天、管理人脉关系、自动回复消息或规划任务提醒
前置准备
- 依赖说明:Python 3.8+,uiautomation库(Windows UI自动化),scikit-learn(文本分析),matplotlib(可视化)
- 系统要求:Windows 10/11,微信桌面版已安装并登录
- 准备工作:
- 确保微信已打开并登录
- 确认要操作的聊天窗口可见
- 准备好知识库文件(kb_data.json)
操作步骤
1. 聊天记录抓取
自动定位并抓取指定聊天窗口的记录:
python scripts/chat_capture.py --contact "联系人名称" --count 100 --output ./chat_records.json
2. 知识库管理
添加对话样本学习用户沟通风格:
python scripts/knowledge_base.py --action add --samples ./samples.json
python scripts/knowledge_base.py --action query --text "需要回复的内容" --top_k 5
3. 需求分析与挖掘
对聊天记录进行深度分析:
python scripts/analyzer.py --input ./chat_records.json --mode analyze --output ./analysis_result.json
4. 自动回复生成
结合知识库生成回复并进入审核流程:
python scripts/auto_reply.py --input "对方消息内容" --contact "联系人" --knowledge_base ./kb_data.json
审核环节会输出候选回复供选择,支持批准/修改/拒绝三种操作。
5. 数据可视化
生成分析图表和人脉关系图:
python scripts/visualizer.py --input ./chat_records.json --type relationship --output ./relationship.png
python scripts/visualizer.py --input ./analysis_result.json --type timeline --output ./timeline.png
使用示例
示例1:抓取并分析聊天记录
- 场景/输入:抓取与"张三"最近200条聊天记录并分析
- 预期产出:JSON格式聊天记录文件 + 分析报告(关键词、频率、情感)
- 关键要点:确保微信窗口可见,联系人名称需精确匹配
示例2:生成并审核回复
- 场景/输入:对方发来"明天下午有空吗?",需要生成回复
- 预期产出:3个候选回复选项 + 审核确认后发送
- 关键要点:先查询知识库学习用户风格,审核后执行发送
示例3:生成人脉关系图
- 场景/输入:分析所有聊天记录生成人脉关系可视化
- 预期产出:PNG格式关系图,节点大小表示互动频率
- 关键要点:需要先抓取足够多的聊天数据
资源索引
- 脚本:见 scripts/chat_capture.py(用途:抓取微信聊天记录,参数:contact/count/output)
- 脚本:见 scripts/knowledge_base.py(用途:知识库管理,参数:action/samples/text/top_k)
- 脚本:见 scripts/analyzer.py(用途:聊天数据分析,参数:input/mode/output)
- 脚本:见 scripts/auto_reply.py(用途:生成回复+审核,参数:input/contact/knowledge_base)
- 脚本:见 scripts/visualizer.py(用途:数据可视化,参数:input/type/output)
- 参考:见 references/format_spec.md(何时读取:定义输入输出数据格式)
注意事项
- 仅在Windows平台使用,需要微信桌面版支持
- 自动发送前必须经过审核确认,不可自动发送
- 抓取频率建议控制在合理范围,避免触发微信限制
- 知识库需要持续更新以提升回复质量
Usage Guidance
Install only if you are comfortable granting local automation access to an already logged-in WeChat desktop session. Use it on chats you are authorized to process, choose narrow contact/count scopes, store exports in a protected folder, delete reply_history.json and kb_data.json when no longer needed, and review every generated reply before any real send action.
Capability Assessment
Purpose & Capability
The stated purpose matches the artifacts: local WeChat chat capture, analysis, knowledge-base matching, visualization, and reviewed reply generation. No network exfiltration, credential theft, or destructive behavior was found.
Instruction Scope
The trigger text covers broad requests like relationship management and reminders, while the actual capability involves WeChat chat capture and possible reply workflows; privacy and consent requirements are not clearly surfaced before ingestion or reply handling.
Install Mechanism
The package declares ordinary Python and Windows UI automation dependencies and contains no installer script, background service, startup persistence, or API-key requirement. Metadata static scan was clean and VirusTotal telemetry was absent.
Credentials
Windows UI automation access to a logged-in WeChat desktop session is purpose-aligned but high impact because it can read private visible chats and interact with the app; the skill relies on user-supplied contact/count commands but does not define strong boundaries for sensitive chats.
Persistence & Privilege
Captured chats, reply requests, reply candidates, and learned conversation samples are written to local JSON files, including user-selected paths and fixed files such as reply_history.json/kb_data.json, without retention, deletion, redaction, permission, or encryption guidance.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install wechat-assistant - After installation, invoke the skill by name or use
/wechat-assistant - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
wechat-assistant 1.0.0 初始版本发布
- 支持自动抓取微信桌面版聊天记录
- 提供聊天记录智能分析与需求挖掘功能
- 集成知识库管理,自定义沟通风格
- 自动生成回复内容,并支持审核发送
- 实现聊天分析和人脉关系等多种数据可视化
- 只支持 Windows 系统和微信桌面版
Metadata
Frequently Asked Questions
What is Wechat Assistant?
微信聊天记录智能分析与自动回复助手;自动抓取聊天记录、智能分析挖掘需求、生成回复内容并支持审核发送;当用户需要分析微信聊天数据、自动生成回复、管理人脉关系或规划任务提醒时使用. It is an AI Agent Skill for Claude Code / OpenClaw, with 39 downloads so far.
How do I install Wechat Assistant?
Run "/install wechat-assistant" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Wechat Assistant free?
Yes, Wechat Assistant is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Wechat Assistant support?
Wechat Assistant is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Wechat Assistant?
It is built and maintained by luemery (@luemery); the current version is v1.0.0.
More Skills