Weave EPUB AI Reader

An integrated Obsidian ebook reader with a built-in DeepSeek reading assistant.

This is an independent GPL-3.0-or-later derivative of Weave EPUB Reader 0.6.55 by Rabbit (zhuzhige). It is not an official release of the upstream project. The exact upstream base is commit 536b2ca29a834385231fe49e6cd757fd07eecd1e.

Why this fork exists

The upstream reader includes an AI button in its native selection toolbar, but the action is delegated to the separate Weave main plugin. This fork keeps the reader's native selected-text pipeline and integrates the AI menu, DeepSeek request, settings, and result window directly into one plugin.

There is no DOM injection, context-menu replacement, runtime patcher, or dependency on the Weave main plugin.

AI reading workflow

1. Select a word, sentence, or paragraph in the reader.
2. Click AI in the reader's native floating selection toolbar.
3. Choose one of five actions:
   • 学术概念解析
   • 英语句子翻译
   • 英语语法解析
   • 通用语境赏析
   • 自定义助手
4. Read or copy the result in a viewport-aware result window.

Every request explicitly includes:

{
  "thinking": {
    "type": "disabled"
  }
}

Requirements

• Obsidian 1.11.4 or newer
• A DeepSeek API key
• Network access to the configured HTTPS API endpoint

The minimum Obsidian version is 1.11.4 because this plugin stores the API key in Obsidian SecretStorage. The key is not written to plugin data.json, source files, build assets, or release archives.

Manual installation

1. Disable Weave EPUB Reader and the old Weave AI Assistant bridge if they are installed. The readers use separate plugin IDs and view types, but Obsidian can associate an ebook file extension with only one reader at a time.

2. Download weave-epub-ai-reader-<version>.zip from the matching GitHub Release and extract it to:

   <Vault>/.obsidian/plugins/weave-epub-ai-reader/
   

3. Confirm that the folder directly contains:

   main.js
   manifest.json
   styles.css
   

4. Reload Obsidian, enable Weave EPUB AI Reader, then open: Settings → Weave EPUB AI Reader → AI Assistant.

5. Save the DeepSeek API key and choose the model, endpoint, maximum output tokens, and optional custom prompt.

The release also contains the three standard Obsidian assets (main.js, manifest.json, and styles.css) for Community Plugins and BRAT. Once the plugin is accepted into the Obsidian community directory, install and update it directly from Settings → Community plugins.

Privacy and network disclosure

• Reading, highlighting, bookshelf state, and normal reader operations are local-first.
• When you explicitly choose an AI action, the selected text, the chosen prompt, model name, and output limit are sent to the configured DeepSeek API endpoint.
• The plugin does not include telemetry.
• The API key is stored through Obsidian SecretStorage in the current vault.
• DeepSeek is a third-party service. Review its privacy policy and terms before using the integration.

See PRIVACY.md for the complete disclosure inherited from and extended beyond the upstream reader.

Building from source

npm ci
npm run build

Release assets are generated in dist/. A valid release contains main.js, manifest.json, styles.css, and an optional convenience ZIP. Never add an API key or a local plugin data.json to a release.

Upstream updates

This repository keeps the upstream Git history. To port a newer reader version:

1. Fetch the upstream tag.
2. Record its tag and exact commit in UPSTREAM.md.
3. Rebase or merge it into a dedicated update branch.
4. Reapply only the integrated-AI source changes if conflicts occur.
5. Run the build, release verification, secret scan, and manual Obsidian selection-toolbar test before publishing.

License and attribution

Licensed under GPL-3.0-or-later. See LICENSE and COPYRIGHT.

• Original reader: zhuzhige123/obsidian-weave-reader
• Integrated AI fork: HarrySuen626 and contributors