WeRead Book Finder

输入图书名称，优先在微信读书搜索并加入书架；若微信读书没有可信命中或无法加入书架，则自动回退到 Z-Library 搜索并优先下载非 PDF 格式。适用于“把《xxx》加入微信读书书架，不行就去 Z-Library 下载”这类请求。

Security Scan

The skill's code and instructions align with its description (attempt WeRead, fall back to Z-Library and download), but it persistently reuses a browser profile and automates downloads from a third‑party site, which has privacy and legal implications the user should consider.

Purpose & Capability

Name/description claim to add a book to WeRead then fall back to Z‑Library; the included Playwright script implements searching, adding to shelf, and falling back to download. Required tools (python, playwright, requests, bs4) match the implementation.

Instruction Scope

SKILL.md and the script instruct the agent/user to log in and reuse a Playwright persistent profile directory so browser sessions are preserved. That scope is consistent with the automation goal but means the skill will read/write browser profile data (cookies, session state) beyond just the target sites if a shared profile is used.

Install Mechanism

No install spec — instruction-only with a Python script. This is low-risk from an installer perspective; dependencies are standard Python and Playwright packages as documented.

Credentials

The skill requests no environment variables or credentials, which is proportional. However, recommending reuse of a browser user_data_dir gives the script persistent access to stored cookies and tokens in that directory; using a shared profile could expose unrelated accounts.

Persistence & Privilege

always:false (normal). The script intentionally persists browser profiles and download files to ~/Downloads/OpenClaw-Books — this is expected for its function but creates long‑lived local state (saved cookies, downloads).

Assessment

This skill appears to do what it says (automate WeRead and, if necessary, search/download from Z‑Library), but consider the following before using it: - Privacy: the tool recommends reusing a Playwright user_data_dir (browser profile). Do NOT point it at a profile that contains other accounts or sensitive sessions; create a dedicated profile directory to avoid exposing unrelated cookies/session tokens. - Downloads and legality: the fallback is Z‑Library, a piracy-associated site. Ensure you understand the legal/privacy implications of downloading content from that source in your jurisdiction. - Sandbox and inspection: run the script in a controlled environment (VM/container) and inspect the code yourself if possible. The code shown performs browser automation and file writes; there are no hidden network endpoints in the provided snippet, but the script will access weread.qq.com and z-lib.by as intended. - Files and storage: downloads go to ~/Downloads/OpenClaw-Books by default; change that path if you want a different location or stricter access controls. - Dependencies: install Playwright and its browser binaries via the documented commands; avoid running arbitrary prebuilt binaries from unknown hosts. If you plan to use this skill, create a fresh browser profile directory just for it, review the full script (including the truncated portion) before running, and run it in an environment where accidental exposure or misuse of other accounts is unlikely.

WeRead Book Finder 自动化

使用 Playwright 脚本自动化处理书籍获取与回退下载。

前提

环境依赖：

• python3
• playwright (安装 Chromium：playwright install chromium)
• requests
• beautifulsoup4

快速使用

指定书名运行：

python3 skills/weread-book-finder/scripts/find_book.py "书名"

登录态管理

由于微信读书和 Z-Library 需要登录，建议固定使用 profile 目录，避免重复登录：

# 登录微信读书
python3 skills/weread-book-finder/scripts/find_book.py --login weread --headed

# 登录 Z-Library
python3 skills/weread-book-finder/scripts/find_book.py --login zlib --headed

后续运行脚本时，会自动复用已登录的浏览器会话。

推荐执行策略

• 默认模式：直接运行脚本，脚本会自动尝试微信读书。若微信读书找不到，脚本会无缝切换至 Z-Library 搜索并下载。
• 严格标题优先：优先精确标题；短标题若只有模糊命中，不会直接误加，而是返回候选后走后备链路。
• 作者辅助匹配：支持输入 书名 / 作者、书名 | 作者、书名 by 作者 这类形式，脚本会把作者纳入排序。
• 候选结果可见：输出里会附带前 5 个候选，便于调试和人工确认。
• 调试模式：添加 --headed 参数可以打开浏览器窗口观察执行过程。
• 文件保存：下载的电子书默认保存在 ~/Downloads/OpenClaw-Books。