teng-lin/weread-omni

What it solves

weread-omni provides an expanded set of capabilities for interacting with WeChat Reading (微信读书) compared to the official Agent Skill. While the official skill is limited to six read-only operations, this project enables over 40 read-write operations, allowing users to manage their bookshelves, modify notes, import personal books, and interact with the platform's AI features programmatically.

How it works

The project functions as an unofficial SDK and CLI tool. It uses a QR-code-based login system to authenticate users. Once authenticated, it provides three primary interfaces for interaction:

• Command Line Interface (CLI): A tool for executing specific reading and writing tasks that can output results in JSON format.
• TypeScript SDK: A fully typed library for integrating WeChat Reading functionality into other applications.
• Agent Skill: A predefined skill set that allows AI agents (like Codex or Claude Code) to check login status, call the CLI, and handle pagination and confirmations for write operations.

It also includes a local content library (powered by SQLite WAL) to cache book information, chapters, and public account articles to reduce redundant network requests.

Who it’s for

• Developers who want to build applications on top of WeChat Reading data.
• Power users who want to automate their reading habits, export public account feeds, or manage their library via the command line.
• AI Agent users who want to give their AI assistants the ability to read and manage their WeChat Reading account.

Highlights

• Extensive Capability: Supports 40+ operations, including adding/editing/deleting bookmarks and reviews, which are not available in the official skill.
• Content Import: Allows importing personal books in EPUB, PDF, MOBI, TXT, and AZW3 formats.
• Public Account Integration: Ability to read, subscribe to, and export feeds from WeChat public accounts.
• AI Integration: Direct access to the "Ask Book" and suggestion features of the WeChat Reading AI.
• Local Caching: Built-in local library to store content and reduce API calls.