wechat daily report

微信群聊天记录日报图片生成工具。分析微信群聊天记录，结合 AI 生成内容，**最终输出为手机端分辨率的日报长图（PNG）**。

Security Scan

The skill's instructions require running analysis/generation scripts and an AI prompt file that are not included in the package, so the delivered bundle is internally inconsistent and incomplete.

Purpose & Capability

The declared purpose (generate a WeChat group daily-report PNG) is plausible and matches the described workflow, but the SKILL.md requires running files (scripts/analyze_chat.py, scripts/generate_report.py, references/ai_prompt.md) that are not present in the package. That mismatch means the package as delivered cannot perform its stated job and may be relying on external code not shown here.

Instruction Scope

Runtime instructions tell the agent/user to analyze chat JSON, produce simplified_chat.txt and stats.json, have an AI produce ai_content.json per references/ai_prompt.md, then render a PNG using Playwright/Chromium. The instructions themselves do not attempt to access unrelated system paths or extra credentials, but they assume missing helper scripts and a missing prompt file — a significant scope/integrity gap. Also the skill processes personal chat logs (sensitive PII) which is expected but should be highlighted to the user.

Install Mechanism

There is no packaged install spec (instruction-only). SKILL.md instructs installing Python deps (jieba, jinja2, playwright) and running 'playwright install chromium' — a reasonable but impactful operation (downloads Chromium). Because no code is installed by the skill itself, the installation risk is limited to the user following those pip commands locally.

Credentials

The skill requests no environment variables, credentials, or config paths. That is proportionate to the stated functionality. However, the skill will process potentially sensitive chat data; the package provides no guidance about safe handling, retention, or where AI generation occurs (local LLM vs external API).

Persistence & Privilege

No 'always' privilege is requested, no installs or persistent background components are bundled. The skill is user-invocable and does not request autonomous/always-on presence.

What to consider before installing

Do not run or trust this package as-is. The SKILL.md references scripts (scripts/analyze_chat.py, scripts/generate_report.py) and references/ai_prompt.md that are missing — you must obtain and review those files before running anything. Because this workflow processes private WeChat group messages, verify the missing scripts do not exfiltrate data, call external endpoints, or upload logs to third parties. If you still want to use it: (1) ask the publisher for the missing scripts and ai_prompt.md and review them line-by-line for network calls and credential usage; (2) run installs (pip, playwright) in an isolated environment or container because playwright will download Chromium; (3) if AI calls are required, confirm which model/service is used and whether data is sent externally; (4) prefer local execution or audit any remote endpoints for TLS and ownership. If the author cannot provide the missing code or clear provenance, treat the skill as incomplete/untrusted.

微信群聊日报生成 Skill

工作流程

1. 运行 analyze_chat.py 分析聊天记录
   ↓
2. AI 根据聊天文本生成内容 (ai_content.json)
   ↓
3. 运行 generate_report.py 生成日报图片 (.png)

⚠️ 最终输出是 PNG 图片，不是 HTML。确保 --output 参数使用 .png 后缀。

使用步骤

1. 分析聊天记录

python scripts/analyze_chat.py <聊天记录.json> --output-stats stats.json --output-text simplified_chat.txt

输出:

• stats.json - 统计数据（话唠榜、熬夜冠军、词云等）
• simplified_chat.txt - 全量聊天文本，供 AI 分析

2. AI 生成内容

根据 simplified_chat.txt 和 stats.json，按 references/ai_prompt.md 中的格式生成 AI 内容，保存为 ai_content.json。

AI 需要生成的内容包括：

• topics: 讨论热点（3-5 个）
• resources: 教程/资源分享
• important_messages: 重要消息
• dialogues: 有趣对话
• qas: 问答
• talker_profiles: 话唠成员的特点标签（常用词已由脚本统计）

3. 生成日报图片

python scripts/generate_report.py --stats stats.json --ai-content ai_content.json --output report.png

✅ 输出后缀必须是 .png，使用 iPhone 14 Pro Max 分辨率 (430x932 @3x)

图片生成需要安装 playwright：pip install playwright && playwright install chromium

聊天记录 JSON 格式

{
  "meta": {
    "name": "群名称",
    "platform": "wechat",
    "type": "group"
  },
  "members": [
    {"platformId": "xxx", "accountName": "昵称"}
  ],
  "messages": [
    {
      "sender": "platformId",
      "accountName": "昵称",
      "timestamp": 1234567890,
      "type": 0,
      "content": "消息内容"
    }
  ]
}

消息类型 (type):

• 0: 纯文本（脚本和 AI 只分析此类型）
• 1: 图片
• 5: 动画表情
• 99: 系统消息

脚本分析内容（保证准确性）

数据	说明
总消息数	所有消息计数
活跃用户数	去重用户数
时间范围	首尾消息时间
话唠榜 TOP3	按发言数排序 + 常用词
熬夜冠军	23:00-06:00 最晚活跃者
词云数据	jieba 分词 + 词频统计

AI 生成内容（需要理解上下文）

内容	输入数据
讨论热点	精简文本 + 词云 TOP50
成员画像	精简文本中的发言
有趣对话	高互动片段
问答识别	问号消息 ± 上下文
教程/资源	精简全文识别

依赖

pip install jieba jinja2 playwright
playwright install chromium