FleurPilot

An Obsidian AI writing companion built for Chinese users — clean UI, zero learning curve, native Chinese experience.

English | 中文

English

Why FleurPilot?

Most Obsidian AI plugins are built with English-first users in mind. FleurPilot flips that: Chinese is the first-class citizen. From the UI to the AI prompts, every detail is tuned for Chinese writing workflows — while still working beautifully in English, Japanese, and Traditional Chinese.

Compared to Copilot-style plugins, FleurPilot is designed around how Chinese users actually work:

• Minimal UI, no clutter — No flashy gradients, no emoji overload, no "AI-looking" interfaces. Clean design inspired by Apple/Linear/Notion aesthetics.
• Chinese-native experience — UI labels, skill names, and AI interactions all feel natural in Chinese. Not a translated afterthought.
• Domestic model support out of the box — DeepSeek, Qwen (通义千问), GLM (智谱), SiliconFlow (硅基流动) are pre-configured. No hunting for API endpoints.
• Inline editing that respects your text — Select, rewrite, preview diff, apply. No copy-paste round-trips to a chat window.
• Zero configuration pressure — Pick your provider, enter the API key, start writing. That's it.

Features

• AI Chat — Side panel with streaming responses, context-aware (current note / folder / all notes)
• Chat / Deep Thinking — Toggle between quick chat and deep reasoning modes
• Inline Editing — Polish, shorten, expand, translate, proofread selected text directly in the editor
• Quick Skills — One-click full-note polish, proofread, and translate (auto-detects Chinese↔English direction)
• Writing Assistant — Full-note review, structure analysis, tone analysis, summary generation
• Right-Click Menu — Quick AI actions on selected text
• Multi-Language UI — Simplified Chinese, Traditional Chinese, English, Japanese

Supported Model Providers

Installation

Option A — Download the bundle (recommended):

1. Download fleurpilot-1.0.2.zip from Releases
2. Extract the zip — you'll get a fleurpilot/ folder containing main.js, styles.css, and manifest.json
3. Move the fleurpilot/ folder into your vault's .obsidian/plugins/ directory
4. Open Obsidian → Settings → Community Plugins → Enable "FleurPilot"
5. Go to plugin settings and enter your API Key

Option B — Manual installation:

1. Download main.js, styles.css, and manifest.json from Releases
2. Create a folder named fleurpilot inside your vault's .obsidian/plugins/ directory
3. Place the three files into that folder
4. Open Obsidian → Settings → Community Plugins → Enable "FleurPilot"
5. Go to plugin settings and enter your API Key

Build from Source

git clone https://github.com/gibsonchan-5/fleurpilot.git
cd fleurpilot
npm install
npm run build

中文

为什么选择 FleurPilot（传彩笔AI）？

市面上的 Obsidian AI 插件几乎都以英文用户为第一优先级，中文体验往往是事后补上的翻译。FleurPilot（传彩笔AI）不同——中文是第一语言。从界面设计到 AI 交互，每个细节都围绕中文写作场景打磨。

与 Copilot 类插件相比，FleurPilot（传彩笔AI）更贴合国内用户的使用习惯：

• 界面极简，拒绝 AI 味 — 没有渐变特效、没有 emoji 堆砌、没有花哨装饰。设计灵感来自 Apple / Linear / Notion 的克制美学，打开即用，无需学习。
• 中文原生体验 — 界面文案、技能名称、AI 交互语言都是地道的中文，不是机翻产物。繁中、英文、日文同样原生支持。
• 国内模型开箱即用 — DeepSeek、通义千问、智谱 GLM、硅基流动已预置 API 地址，用户只需填入 Key 即可开始使用，不用自己查找接口文档。
• 侵入式编辑，所见即所得 — 选中文本 → 右键改写 → 预览 diff → 一键应用。不用在聊天窗口和编辑器之间来回复制粘贴。
• 零配置压力 — 选模型、填 Key、开始写。三步搞定，没有复杂的参数面板。

功能亮点

• AI 对话 — 侧边栏流式对话，支持携带当前笔记 / 文件夹 / 全部笔记作为上下文
• Chat / 深度思考 — 一键切换普通对话与深度推理模式
• 侵入式编辑 — 选中文本直接改写（精简、扩写、润色、翻译、校对），编辑器内预览 diff 并应用
• 快捷技能 — 一键全文润色、智能校对、全文翻译（自动识别中英方向）
• 写作助手 — 全文审读、结构分析、风格分析、内容摘要
• 右键菜单 — 编辑器中选中文本即可快速调用 AI
• 多语言界面 — 简体中文、繁体中文、英文、日文

支持模型

安装

方式 A — 一键打包下载（推荐）：

1. 从 Releases 下载 fleurpilot-1.0.2.zip
2. 解压后会得到一个 fleurpilot/ 文件夹，里面包含 main.js、styles.css、manifest.json 三个文件
3. 把 fleurpilot/ 文件夹整个移动到 vault 的 .obsidian/plugins/ 目录下
4. 打开 Obsidian → 设置 → 社区插件 → 启用「FleurPilot（传彩笔AI）」
5. 进入插件设置，填入 API Key

方式 B — 手动安装：

1. 从 Releases 下载 main.js、styles.css、manifest.json 三个文件
2. 在 vault 的 .obsidian/plugins/ 目录下创建 fleurpilot 文件夹
3. 将三个文件放入该文件夹
4. 打开 Obsidian → 设置 → 社区插件 → 启用「FleurPilot（传彩笔AI）」
5. 进入插件设置，填入 API Key

从源码构建

git clone https://github.com/gibsonchan-5/fleurpilot.git
cd fleurpilot
npm install
npm run build

⚠️ Security

API Key is stored locally in your vault's plugin data file (data.json). Please keep it safe and do not share with others.

Note content (including the current note / selected folder notes used as context) is sent only to the LLM API endpoint you configure (e.g. DeepSeek, Qwen, GLM, SiliconFlow, or a custom endpoint). No telemetry, analytics, or third-party tracking is included.

API Key 存储在 vault 的插件数据文件（data.json）中，请妥善保管，勿与他人分享。

笔记内容（含作为上下文的当前笔记 / 所选文件夹笔记）仅发送至你在设置中配置的 LLM API 地址（如 DeepSeek、通义千问、智谱、硅基流动或自定义端点）。本插件不含任何遥测、统计或第三方追踪代码。

License

MIT License. See LICENSE for details.