← Back to Skills Marketplace
lwter

读书每日推荐

by Neway Lau · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ Security Clean
43
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install nexus-reader
Description
微信读书飙升榜每日推荐卡片生成器。从微信读书飙升榜抓取热门书籍数据,生成精美的每日读书推荐卡片(HTML/PNG)。当用户说「读书推荐」「微信读书」「飙升榜」「今日好书」「推荐一本书」「读书卡片」「book recommendation」时触发。也可用于每日定时推送读书推荐场景。
README (SKILL.md)

nexus-reader - 微信读书飙升榜每日推荐

生成精美的每日读书推荐卡片,数据来源于微信读书飙升榜

核心工作流程(四步)

Step 1:抓取榜单数据

python scripts/reader.py --fetch-only

脚本会自动:

  1. 抓取微信读书飙升榜(20本书)
  2. 保存原始数据到 data/weread_rising_YYYY-MM-DD.json
  3. 输出榜单摘要(书名、作者、推荐值)

输出示例:

[抓取] 正在获取微信读书飙升榜...
[解析] 找到 20 个书籍项
[保存] 数据已缓存: data/weread_rising_2026-05-04.json

==================================================
飙升榜 TOP 20
==================================================
  1. 《阿勒泰的角落》- 李娟 (推荐值: 88.4%)
  2. ...

缓存机制:当天JSON文件 \x3C 1小时自动跳过重新抓取,--force 强制刷新


Step 2:模型选书 + 创作内容

读取 data/weread_rising_YYYY-MM-DD.json,从TOP10中选一本书,然后自己创作以下内容(不要直接复制书籍简介):

字段 说明 字数建议
title 金句标题,根据书籍主题提炼一句有深度、有力量的话 8-15字
yi 宜××,结合书籍主题创作 4-6字
ji 忌××,结合书籍主题创作 4-6字
book_desc 推荐语,用温暖、有感染力的语言为这本书写一段推荐 50-100字

示例(推荐《百年孤独》):

{
  "title": "每个人都在守护自己的孤独",
  "yi": "静心阅读",
  "ji": "浮躁焦虑",
  "book_desc": "马尔克斯用一座虚构的小镇,写尽了人类共有的孤独与执念。每一次重读,都会在某个角色身上看见自己。适合在一个安静的下午,泡一杯咖啡,慢慢走进马孔多。"
}

重要book_desc 是 LLM 的创作内容,不是书籍简介的截取!


Step 3:生成 card JSON 文件

将以下内容保存为 data/nexus-reader-YYYYMMDD.json

{
  "date": "2026/05/04",
  "day": "04",
  "weekday": "星期日",
  "lunar_date": "农历 四月十八",
  "yi": "静心品读",
  "ji": "焦虑内耗",
  "title": "有一分热就去发光",
  "book_title": "阿勒泰的角落",
  "book_author": "李娟",
  "book_desc": "...(LLM创作的推荐语)",
  "cover_url": "https://weread-1258476245.cos.ap-nanjing.myqcloud.com/...",
  "book_url": "https://weread.qq.com/web/reader/...",
  "rating": "88.4%",
  "reading_count": "1486"
}

提示cover_urlbook_urlratingreading_countbook_titlebook_authorweread_rising_YYYY-MM-DD.json 中对应的书籍项获取。其余字段由 LLM 创作或计算。


Step 4:渲染卡片

# 只生成 HTML
python scripts/reader.py --card data/nexus-reader-2026-05-04.json

# 生成 HTML + PNG 图片
python scripts/reader.py --card data/nexus-reader-2026-05-04.json --png

输出:

  • data/nexus-reader-YYYYMMDD.html — 精美推荐卡片
  • data/nexus-reader-YYYYMMDD.png — PNG 图片(需加 --png

然后用 preview_url 打开 HTML 卡片展示给用户。


命令行参数速查

命令 说明
--fetch-only 阶段1:只抓取数据
--card JSON_PATH 阶段2:渲染指定 JSON 的卡片
--png 额外生成 PNG(需配合 --card
--output DIR 指定输出目录
--force 强制重新抓取
--cache-hours N 缓存有效时间(小时,默认1)

--fetch-only--card 互斥,必须指定其中一个


数据结构

原始榜单 JSON(weread_rising_YYYY-MM-DD.json

从微信读书抓取的原始数据,包含20本书:

[
  {
    "index": "1",
    "title": "阿勒泰的角落",
    "author": "李娟",
    "desc": "书籍简介...",
    "cover_url": "https://...",
    "book_url": "https://weread.qq.com/...",
    "reading_count": "1486",
    "rating": "88.4%",
    "date": "2026-05-04"
  }
]

卡片 JSON(nexus-reader-YYYYMMDD.json

由 LLM 创作内容后生成的完整卡片数据,用于渲染:

{
  "date": "2026/05/04",
  "day": "04",
  "weekday": "星期日",
  "lunar_date": "农历 四月十八",
  "yi": "静心品读",
  "ji": "焦虑内耗",
  "title": "有一分热就去发光",
  "book_title": "阿勒泰的角落",
  "book_author": "李娟",
  "book_desc": "李娟用质朴的笔触...",
  "cover_url": "https://...",
  "book_url": "https://...",
  "rating": "88.4%",
  "reading_count": "1486"
}

卡片设计说明

卡片采用暖色调书卷风格:

  • 背景:#fdfcfa(米白色)+ 淡雅渐变
  • 主色:#9d7a58(古铜色)用于日期
  • 字体:Noto Serif SC(宋体风格衬线字体)
  • 结构:日期 → 宜忌 → 金句标题+推荐语+书名 → 分隔线 → 封面+详情+阅读按钮

农历日期计算

lunar_date 字段由 LLM 根据当天日期计算,格式为 农历 X月XX日

示例中 LUNAR_MONTHS 为简化版本,实际农历需要专用库(lunarcalendar)或 LLM 直接计算。推荐让 LLM 直接计算,避免额外依赖。


注意事项

  • 微信读书页面不需要登录即可查看飙升榜
  • 封面图片来自 CDN,可能在某些网络环境下无法加载
  • --png 需要安装 html2imagepip install html2image
  • 推荐值和阅读人数是实时数据,每天不同
  • 如果抓取失败,脚本会尝试使用旧缓存降级

定时推送场景

当用户要求每日定时推送读书推荐时,创建 automation(recurring):

  1. prompt 内容:运行 reader.py --fetch-only → LLM 创作内容 → 生成 card JSON → 运行 reader.py --card [--png] → 推送
  2. schedule 为每日固定时间(如 RRULE:FREQ=DAILY;BYHOUR=8;BYMINUTE=0
Usage Guidance
This skill is reasonable for generating book recommendation cards from public WeRead data. Before installing, check whether you are comfortable running the local Python helper, installing optional PNG dependencies, opening generated HTML, and enabling any daily recurring push.
Capability Analysis
Type: OpenClaw Skill Name: nexus-reader Version: 1.0.0 The nexus-reader skill is a legitimate tool for scraping book rankings from WeChat Reading and generating HTML/PNG recommendation cards. The Python script (scripts/reader.py) performs standard web scraping and file operations within a local data directory, and the instructions in SKILL.md clearly define a benign workflow for content creation and rendering without any evidence of malicious intent or data exfiltration.
Capability Assessment
Purpose & Capability
The artifacts are coherent with the stated purpose: they fetch public WeRead rising-list data, cache it locally, and render HTML/PNG recommendation cards.
Instruction Scope
The skill instructs the agent to run local Python commands and can create a recurring automation when the user asks for daily push; these are disclosed and purpose-aligned.
Install Mechanism
There is no install spec, but the code uses Python libraries and optional browser-based PNG rendering; users may need to review and install dependencies manually.
Credentials
Network access to WeRead/CDNs and local writes under a data directory are expected for this card generator. Generated HTML should be treated carefully if supplied with untrusted JSON.
Persistence & Privilege
No credentials or elevated privileges are requested, but the optional daily automation is persistent and should only be enabled with an intended schedule and destination.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install nexus-reader
  3. After installation, invoke the skill by name or use /nexus-reader
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
nexus-reader v1.0.0 - 初始发布,支持从微信读书飙升榜每日抓取热门书籍数据 - 自动生成含金句标题、宜忌、个性化推荐语的精美每日读书推荐卡片(HTML/PNG) - 内置缓存、榜单解析、LLM内容创作、农历日期计算等核心流程 - 支持命令行控制抓取、生成卡片及图片输出 - 可用于关键词触发或每日定时自动推送推荐
Metadata
Slug nexus-reader
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is 读书每日推荐?

微信读书飙升榜每日推荐卡片生成器。从微信读书飙升榜抓取热门书籍数据,生成精美的每日读书推荐卡片(HTML/PNG)。当用户说「读书推荐」「微信读书」「飙升榜」「今日好书」「推荐一本书」「读书卡片」「book recommendation」时触发。也可用于每日定时推送读书推荐场景。 It is an AI Agent Skill for Claude Code / OpenClaw, with 43 downloads so far.

How do I install 读书每日推荐?

Run "/install nexus-reader" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is 读书每日推荐 free?

Yes, 读书每日推荐 is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does 读书每日推荐 support?

读书每日推荐 is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created 读书每日推荐?

It is built and maintained by Neway Lau (@lwter); the current version is v1.0.0.

💬 Comments