SuperAPI AI Directory
Partner with us中文AI newsFavorites
Skillswechat-daily-report

wechat-daily-report

Data & analytics

View repository

Updated 10-16 05:06

About this skill

Objective

Read WeChat chat history from local DuckDB and generate a structured today's report Markdown file, including:

  • Data overview (total messages, active chats, time distribution)
  • To-do items: Categorized by priority, summarizing and extracting work that needs follow-up
  • Fun gossip: Organized into complete stories, rather than fragmented messages
  • End with a humorous summary
Process

Step 1: Extract raw data

Run the script to get the day's chat data:

python3 scripts/extract_data.py [--date YYYY-MM-DD] [--db PATH]

The script will output:

  1. Data statistics (message count, activity level, time distribution)
  2. Complete work group conversations (for analyzing to-dos)
  3. Non-work group conversations (for extracting fun stories)

Step 2: Summarize and extract (done by AI)

Based on the raw data, analyze and summarize:

To-do item extraction rules

  1. Identify to-do sources：

    • Messages where @淇奥 is mentioned
    • Conversations containing questions/faults/requests
    • Items that need confirmation/follow-up/feedback

  2. Categorize by priority：

    • 🔴 Needs immediate handling: things to do today, things already promised
    • 🟡 Follow up this week: those with clear time nodes
    • 🟢 Resolved/Noted: completed or only need to be aware of

  3. Format requirements：

    1. **事项标题**
   - 来源：群名/联系人 + 时间
   - 具体内容描述
   - 状态：当前进展

Fun gossip extraction rules

  1. Identify interesting content：

    • Topics hotly discussed by group members (multiple people participating in discussion)
    • Contains gossip/tea/comedic keywords
    • Conversations with a complete storyline

  2. Organize into stories：

    • Give each story a title + emoji
    • Summarize the event process, retain brilliant original quotes
    • Highlight the punchlines/roast points

  3. Format requirements：

    ### 🍉 故事标题

故事概括描述，**加粗关键信息**，保留群友精彩原话用引用格式。

Today's summary

End with a humorous sentence, combining the day's to-dos and fun stories, for example:

33 work groups battling online, the Git folder still hasn't been created, but the most important thing is—the red packet for playing Jianwang 3 cannot be claimed.

Step 3: Generate the report

Organize the analysis results into a complete Markdown report and save it to the specified directory.

Report template 
# 微信日报 (YYYY-MM-DD 周X)

## 一、数据概览
| 指标 | 数值 |
|------|------|
| **消息总数** | X 条 |
| **活跃聊天** | X 个 |
| **活跃人数** | X 人 |
| **峰值时段** | XX:00 (X条) |

## 二、聊天活跃度 TOP 10
（表格）

## 三、消息类型分布
（列表）

## 四、待办事项
### 🔴 需立即处理
### 🟡 本周跟进
### 🟢 已解决/知悉

## 五、趣事吃瓜
### 🍉 故事1
### 📱 故事2
...

## 六、活跃发送者 TOP 10
（表格）

## 七、消息时间分布
（柱状图）

## 今日总结
> 幽默收尾语

---
*报告生成时间: YYYY-MM-DD HH:MM:SS*
Guardrails
  • Read-only operations, do not modify the database
  • To-do items are extracted only from work-related chats
  • Fun gossip does not include sensitive/private content
  • The report is output to the current directory by default
Resources
  • scripts/extract_data.py: Data extraction script
  • scripts/generate_report.py: Basic report generation script (statistics part)
Usage example 
# 1. 提取今日数据
python3 scripts/extract_data.py

# 2. AI 根据输出概括提炼，生成最终报告

# 或者直接让 AI 生成（推荐）
# "帮我生成今日微信报告"