HeriHeriCloud, 一个更现代、更强大的蓝奏云封装。无限存储 + 不限速,支持 WebDAV,MCP 与高级网盘功能,让蓝奏云真正好用。
개요
HeriHeriCloud 是一个跨平台的云盘客户端解决方案。本项目以蓝奏云的无限存储空间和不限速带宽为底层基础设施,新增WebDAV挂载、AI智能体接入(MCP)与其他高级功能,旨在为用户提供现代化、高性能的文件管理与分享体验。 当前市场上,并没有哪一款云盘可以做到真正的不限速不限存储空间,包括蓝奏云。不过,各个主流云盘变得更加臃肿是不争的事实。它们为了塞进去更多的功能,从而抛弃了一个云盘最原始的作用:存储,下载以及浏览。同时,各个网盘开始封闭自己的环境生态,同时也将市场挤兑到饱和。我们对自己的定位很清晰,蓝奏云对于下载不限速,对于上传的限制普通用户也很难触及,同时也是做到不限存储空间,所以HeriHeriCloud旨在给用户提供一个简单便捷高效的云盘体验。用户可以只是用来传输一些资料而不需要去开通其他网盘的会员,用户也可以是希望搭建自己的小小的媒体库用WebDAV挂载观看视频电影,我们希望能覆盖普通用户到有专业网盘需求的用户。关于WebDAV挂载,目前多端设备均支持作为WebDAV代理,可以在局域网内挂载到Infuse,网易爆米花等媒体中心,同时也支持Raidrive挂载至本地网盘。对于HeriHeriCloud的网盘资源,目前仅限于用户自己上传使用,如果有人愿意在互联网上分享HeriHeriCloud网盘以及自己的资源链接,我们会无比荣幸。当然,目前来讲,一些蓝奏云盘的限制是无法避免的,比如说蓝奏云会限制每24小时中用户上传的流量大小,尚不明确具体的数值,但是可以确定这个限制是存在的。 本项目原生支持大语言模型智能体(如 OpenCode, CC Switch, Cursor, Claude Desktop 等)进行自动化文件操作! * :在项目根目录的 skill/ 文件夹中可以找到预打包的技能 ZIP 文件。在您的智能体客户端技能面板中,选择“从 ZIP 安装技能”,即可让 AI 自动为您上传、下载和管理云盘文件(AI 操作将被严格限制在安全的 AgentWorkspace 沙盒文件夹内)。 * :支持标准 STDIO 模式,在您的 MCP 配置文件中添加启动参数 --agent(如 heriheri.exe --agent),即可在后台开启无头模式对接。
README
English | 中文
HeriHeriCloud
HeriHeriCloud 是一个跨平台的云盘客户端解决方案。本项目以蓝奏云的无限存储空间和不限速带宽为底层基础设施,新增WebDAV挂载、AI智能体接入(MCP)与其他高级功能,旨在为用户提供现代化、高性能的文件管理与分享体验。
当前市场上,并没有哪一款云盘可以做到真正的不限速不限存储空间,包括蓝奏云。不过,各个主流云盘变得更加臃肿是不争的事实。它们为了塞进去更多的功能,从而抛弃了一个云盘最原始的作用:存储,下载以及浏览。同时,各个网盘开始封闭自己的环境生态,同时也将市场挤兑到饱和。我们对自己的定位很清晰,蓝奏云对于下载不限速,对于上传的限制普通用户也很难触及,同时也是做到不限存储空间,所以HeriHeriCloud旨在给用户提供一个简单便捷高效的云盘体验。用户可以只是用来传输一些资料而不需要去开通其他网盘的会员,用户也可以是希望搭建自己的小小的媒体库用WebDAV挂载观看视频电影,我们希望能覆盖普通用户到有专业网盘需求的用户。关于WebDAV挂载,目前多端设备均支持作为WebDAV代理,可以在局域网内挂载到Infuse,网易爆米花等媒体中心,同时也支持Raidrive挂载至本地网盘。对于HeriHeriCloud的网盘资源,目前仅限于用户自己上传使用,如果有人愿意在互联网上分享HeriHeriCloud网盘以及自己的资源链接,我们会无比荣幸。当然,目前来讲,一些蓝奏云盘的限制是无法避免的,比如说蓝奏云会限制每24小时中用户上传的流量大小,尚不明确具体的数值,但是可以确定这个限制是存在的。
AI 智能体 / MCP 接入 (全新功能)
本项目原生支持大语言模型智能体(如 OpenCode, CC Switch, Cursor, Claude Desktop 等)进行自动化文件操作!
- OpenCode / CC Switch 用户:在项目根目录的
skill/文件夹中可以找到预打包的技能 ZIP 文件。在您的智能体客户端技能面板中,选择“从 ZIP 安装技能”,即可让 AI 自动为您上传、下载和管理云盘文件(AI 操作将被严格限制在安全的AgentWorkspace沙盒文件夹内)。 - Cursor / Claude Desktop 用户:支持标准 STDIO 模式,在您的 MCP 配置文件中添加启动参数
--agent(如heriheri.exe --agent),即可在后台开启无头模式对接。 - REST API 模式:正常打开 HeriHeriCloud 客户端后,MCP 服务端会默认运行在
http://127.0.0.1:8888/mcp,支持自定义智能体通过 HTTP POST 调用。
OpenWRT 路由器支持 (全新功能)
本项目提供超轻量的无头(Headless)守护进程,支持在多种架构的 OpenWRT 路由器上常驻运行,将您的路由器无缝打造成局域网流媒体与文件共享中心!
- 获取与运行:在 GitHub Releases 页面下载对应路由器架构的单文件程序(支持
aarch64、armv7、mipsel)。通过 SSH/SCP 上传至路由器,执行chmod +x heriheri-openwrt-*赋予权限后直接运行即可。 - 后台与开机自启:由于本程序为纯静态编译的单文件(非 ipk),若要使其在关闭 SSH 后保持后台运行,请使用命令
nohup /存放路径/heriheri-openwrt-* >/dev/null 2>&1 &。如需开机自动运行,请在 OpenWRT Web 后台的“系统 -> 启动项 -> 本地启动脚本”(即/etc/rc.local)中,将该启动命令添加至exit 0上方。 - 内置 WebUI 配置:程序启动后,在浏览器中访问
http://:8888即可打开内置的极简可视化面板,完成蓝奏云账号登录、强制同步 VFS 以及修改 WebDAV 端口和账密。 - 局域网原画串流:配置完成后,同局域网下的智能电视、Apple TV (Infuse)、本地电脑等均可通过 WebDAV 协议直接挂载,免下载直接观看云盘内的原画视频,体验极速局域网串流。
核心特性
- 跨平台架构:基于 Tauri v2 + TypeScript + React 构建,确保在提供原生级性能和极低内存占用的同时,实现全平台代码复用。
- 原生 WebDAV 服务:内置轻量级 WebDAV 代理,将深层目录结构和文件直接映射为标准 WebDAV 协议,支持第三方播放器串流与系统级挂载。
- AI 智能体支持 (MCP):内置 Model Context Protocol (MCP) 双模式服务器,无缝接入 Cursor、OpenCode 等各大主流 AI 智能体,允许 AI 在安全沙盒内自主检索、上传、下载与管理云端文件。
- OpenWRT 路由器支持:提供超轻量的无头(Headless)守护进程,支持在多种架构(ARM/MIPS/x86)的 OpenWRT 路由器上 24/7 常驻运行,配合内置的单页 WebUI 配置面板,轻松将路由器打造成局域网 WebDAV 媒体流转中心。
- 自定义虚拟文件系统 (VFS):突破底层云盘的目录层级和文件类型限制,在本地构建并维护完整的树状目录结构。
- 高级媒体与文档预览:内置视频流媒体实时缓冲代理,并支持文本、代码、图片及 PDF 文档的原生窗口预览。
- 极速并发传输:支持自定义并发数与速度限制,充分利用带宽上限,实现大文件的上传下载以及断点续传。
- 存储可靠:账户注册与登录完全使用蓝奏云盘账户,不存在第三方泄露。同时后端存储直接为用户自己的蓝奏云盘,不存在文件丢失之类的情况。
构建与运行
环境准备
确保您的计算机已安装以下环境:
- Node.js (建议 v20+)
- Rust 工具链 (rustup, cargo)
- Tauri V2 开发依赖 (如 macOS 的 Xcode 命令行工具,Linux 的 webkit2gtk 等)
编译步骤
- 克隆项目仓库:
git clone https://github.com/AgainsTurb/HeriHeriCloud.git
cd HeriHeriCloud
- 安装前端依赖:
npm install
- 关键配置:出于安全和加解密机制的需求,编译器在构建时需要读取一个环境变量。您必须在项目根目录或全局创建
.cargo/config.toml文件,并提供一个测试密钥:
创建或编辑 .cargo/config.toml,填入以下内容:
[env]
HERIHERI_SECRET_KEY = "dummy_secret_key"
CHUNKS_NAMES_KEY = "dummy_secret_key"
注意,自行创建的测试密钥仅限本地测试功能,与Release版本不兼容。
- 启动开发环境:
- Windows, macOS, Linus
npm run tauri dev
- Android
npm run tauri android dev
- iOS
npm run tauri ios dev
- 编译生产版本:
- Windows, macOS, Linus
npm run tauri build
- Android
npm run tauri android build --apk
- iOS
npm run tauri ios build
贡献指南
我们非常欢迎来自社区的贡献,无论是功能开发、Bug 修复还是界面优化。
代码规范
为了保持代码库的统一性和可维护性,所有提交的代码中,变量命名、函数命名以及代码注释必须全部使用英文。
交流与反馈
GitHub 的 Issues 和 Pull Requests 讨论区支持使用任何语言(中文或英文皆可)。
Fork 与项目搭建
-
点击项目页面右上角的 “Fork” 按钮,将仓库 Fork 到你的 GitHub 账户下。
-
将你 Fork 的仓库克隆到本地:
git clone https://github.com/你的用户名/仓库名.git
- 进入项目目录:
cd 仓库名
- 添加原始仓库为 upstream 远程仓库,以便同步更新:
git remote add upstream https://github.com/AgainsTurb/HeriHeriCloud.git
创建分支
每次开发前请新建一个分支,并使用清晰的分支名:
git checkout -b feature/你的功能名称
进行修改
-
请严格遵守上述代码规范。
-
保持修改聚焦,并写好英文注释。
-
在本地充分测试你的改动。
提交与推送
- 暂存修改的文件:
git add .
- 使用清晰简洁的英文提交信息进行提交:
git commit -m "Fix errors"
- 将分支推送到你 Fork 的仓库:
git push origin feature/你的功能名称
发起 Pull Request
-
访问你 Fork 后的 GitHub 仓库页面,通常会看到一个提示你发起 Pull Request 的横幅,点击 “Compare & pull request”。
-
确保 base 分支选择原始仓库的
main(或master),compare 分支选择你的功能分支。 -
填写清晰的标题和详细的改动说明,如有相关 Issue 也请提及。
-
提交 Pull Request。维护者会进行审查,可能需要你进一步修改。
发布流程(标签)
如果你是维护者,需要发布新版本:
- 在完成改动后,暂存并提交(不要忘记修改 CHANGELOG.md):
git add .
git commit -m "Fix errors"
- 使用 npm 更新版本号:
npm version x.x.x
(npm version patch) (npm version minor) (npm version major)
此命令会自动更新 package.json、创建版本提交和对应的标签。
- 推送提交和标签:
git push
git push origin vx.x.x
API 声明
本项目中涉及蓝奏云的所有底层 API 交互逻辑、协议解析及逆向工程工作,均由作者从零开始独立研究并编写,未参考、借用或依赖任何其他的第三方开源项目或现有实现。所有接口逻辑均保持绝对最新,因此本项目不涉及对其他非基础库项目的致谢声明。
结语
真诚欢迎各位提交 Issue 报告问题或发起 PR 改进代码。我们鼓励大家分发、打包并传播这个应用程序,共同建立一个开放的资源共享社区。
如果这个项目对您有帮助,请考虑留下一个 Star🌟。
HeriHeriCloud
HeriHeriCloud is a cross-platform cloud storage client solution. Built upon the infrastructure of Lanzou Cloud—leveraging its unlimited storage and unthrottled bandwidth—it incorporates WebDAV mounting, AI agent integration (MCP), and other advanced features to provide users with a modern, high-performance file management and sharing experience.
Currently, no cloud storage service on the market offers truly unlimited speed and storage space—not even Lanzou Cloud. However, it is an undeniable fact that mainstream cloud services have become increasingly bloated; in their rush to cram in extra features, they have lost sight of the core functions of cloud storage: uploading, downloading, and browsing. Meanwhile, these services have created closed ecosystems and saturated the market. We have a clear vision for ourselves: HeriHeriCloud offers unlimited download speeds, upload limits that are virtually unreachable for the average user, and unlimited storage space. Our goal is to provide a simple, convenient, and efficient cloud storage experience. Whether a user simply needs to transfer files without purchasing a premium subscription elsewhere, or wants to build a personal media library—using WebDAV to stream movies and videos—we aim to cater to everyone from casual users to those with professional storage needs. Regarding WebDAV support, our service allows various devices to act as WebDAV proxies; they can be mounted within a local network to media centers like Infuse or NetEase Popcorn, or mounted as a local drive using RaiDrive. Currently, HeriHeriCloud resources are intended for personal use by the uploader, though we would be honored if users chose to share their HeriHeriCloud links publicly. Of course, certain limitations inherent to the underlying Lanzou Cloud infrastructure are unavoidable; for instance, there is a cap on the total volume of data a user can upload within a 24-hour period. While the exact figure is unspecified, the restriction is certainly in effect.
AI Agent / MCP Integration (NEW)
This project natively supports Large Language Model agents (like OpenCode, CC Switch, Cursor, Claude Desktop, etc.) for automated file management!
- OpenCode / CC Switch Users: You can find the pre-packaged skill ZIP file in the
skill/folder at the project root. In your agent client’s skills panel, select “Install from ZIP” to empower your AI to upload, download, and manage your cloud drive files automatically (all AI operations are strictly sandboxed within theAgentWorkspacefolder). - Cursor / Claude Desktop Users: We support the standard STDIO MCP mode. Simply add the
--agentlaunch flag in your MCP configuration (e.g.,heriheri.exe --agent) for a headless background connection. - REST API Mode: When the HeriHeriCloud client is running normally, the MCP server actively listens at
http://127.0.0.1:8888/mcp, allowing custom agents to connect via HTTP POST requests.
OpenWRT Router Support (NEW)
This project provides an ultra-lightweight headless daemon designed to run continuously on multi-architecture OpenWRT routers, seamlessly transforming your router into a local network media and file sharing hub!
- Download & Run: Download the standalone binary for your specific architecture (
aarch64,armv7,mipsel) from the GitHub Releases page. Upload it to your router via SSH/SCP, grant execution permissions withchmod +x heriheri-openwrt-*, and run it directly. - Persistence & Auto-Start: Because this is a statically compiled standalone binary (not an
.ipk), running it directly ties it to your SSH session. To keep it running in the background, start it withnohup /path/to/heriheri-openwrt-* >/dev/null 2>&1 &. For auto-start on router reboot, navigate to System -> Startup -> Local Startup in your OpenWRT LuCI WebGUI (which edits/etc/rc.local) and paste thenohupcommand just aboveexit 0. - Built-in WebUI: Once running, access
http://:8888in your browser to open the minimal configuration panel. Here, you can log into your Lanzou Cloud account, force a VFS sync, and customize your WebDAV port and credentials. - Local Network Streaming: After configuration, any device on the same local network (e.g., Apple TV with Infuse, Smart TVs, PCs) can mount the drive via the WebDAV protocol to stream original-quality cloud videos directly, enjoying blazing-fast local network speeds without downloading.
Key Features
- Cross-Platform Architecture: Built with Tauri v2 + TypeScript + React, ensuring native-level performance and exceptionally low memory usage while achieving code reuse across all platforms.
- Native WebDAV Service: Features a built-in lightweight WebDAV proxy that maps deep directory structures and files into the standard WebDAV protocol, supporting third-party player streaming and system-level mounting.
- OpenWRT Router Support: Provides an ultra-lightweight headless daemon designed to run 24/7 on multi-architecture (ARM/MIPS/x86) OpenWRT routers. It includes a built-in single-page WebUI to effortlessly transform your router into a local network WebDAV media streaming hub.
- AI Agent Support (MCP): Features a built-in Model Context Protocol (MCP) dual-mode server, seamlessly connecting with major AI agents like Cursor and OpenCode. This allows AI to autonomously search, upload, download, and manage your cloud assets within a secure sandbox.
- Virtual File System (VFS): Overcomes the directory depth and file type restrictions of the underlying cloud drive by building and maintaining a complete, local tree-based directory structure.
- Advanced Media & Document Preview: Includes a real-time buffering proxy for video streaming, along with native window previewers for text, code, images, and PDF documents.
- High-Speed Concurrent Transfers: Supports custom concurrency limits and speed throttling, fully utilizing bandwidth capacity for large file chunk uploads, downloads and resumable transfer.
- Reliable Storage: Account registration and login rely entirely on Lanzou Cloud accounts, eliminating the risk of third-party data leaks. Furthermore, since data is stored directly in the user’s own Lanzou Cloud account, there is no risk of file loss.
Build and Run
Prerequisites
Ensure your machine has the following installed:
- Node.js (v20+ recommended)
- Rust Toolchain (rustup, cargo)
- Tauri V2 system dependencies (e.g., Xcode Command Line Tools for macOS, webkit2gtk for Linux)
Build Instructions
- Clone the repository:
git clone [https://github.com/yourusername/HeriHeriCloud.git](https://github.com/yourusername/HeriHeriCloud.git)
cd HeriHeriCloud
- Install frontend dependencies:
npm install
- Critical Configuration: Due to security and encryption mechanisms, the compiler requires a specific environment variable during the build process. You must create a
.cargo/config.tomlfile in the project root or your global cargo directory and provide a dummy key:
Create or edit .cargo/config.toml and add the following:
[env]
HERIHERI_SECRET_KEY = "dummy_secret_key"
CHUNKS_NAMES_KEY = "dummy_secret_key"
Note that self-generated test keys are intended solely for local testing and are incompatible with the release version.
- Run the development environment:
- Windows, macOS, Linus
npm run tauri dev
- Android
npm run tauri android dev
- iOS
npm run tauri ios dev
- Build the production release:
- Windows, macOS, Linus
npm run tauri build
- Android
npm run tauri android build --apk
- iOS
npm run tauri ios build
Contributing
We warmly welcome contributions from the community, whether it’s feature development, bug fixing, or UI improvements.
Code Standards
To maintain consistency and maintainability across the codebase, all variable names, function names, and code comments within submitted code must be entirely in English.
Communication
Discussions in GitHub Issues and Pull Requests are welcome in any language (Chinese or English).
Fork and Set Up
-
Fork the repository by clicking the “Fork” button at the top right of the project page.
-
Clone your fork locally:
git clone https://github.com/YOUR-USERNAME/REPOSITORY-NAME.git
- Navigate into the project directory:
cd REPOSITORY-NAME
- Add the original repository as an upstream remote to keep your fork in sync:
git remote add upstream https://github.com/AgainsTurb/HeriHeriCloud.git
Create a Branch
Always create a new branch for your work. Choose a descriptive branch name:
git checkout -b feature/my-awesome-feature
Make Your Changes
-
Follow the code standards above.
-
Keep changes focused and well‑documented (comments in English).
-
Test your changes locally.
Commit and Push
- Stage your changes:
git add .
- Commit with a clear, concise message (English is preferred for commit messages):
git commit -m "Fix errors"
- Push your branch to your fork:
git push origin feature/my-awesome-feature
Open a Pull Request
-
Visit your fork on GitHub. You will see a banner prompting you to open a Pull Request. Click “Compare & pull request”.
-
Ensure the base branch is the original repository’s
main(ormaster) and the compare branch is your feature branch. -
Write a descriptive title and a clear explanation of your changes. Mention any related issues.
-
Submit the Pull Request. The maintainers will review it and may request further adjustments.
Release Process (Tags)
If you are a maintainer creating a new release:
- After finalising your changes, stage and commit them (Do not forget to update CHANGELOG.md):
git add .
git commit -m "Fix errors"
- Bump the version using npm:
npm version vx.x.x
(npm version patch) (npm version minor) (npm version major)
This updates package.json and creates a version commit and tag automatically.
- Push the commit and the tag:
git push
git push origin vx.x.x
API Statement
All underlying API interaction logic, protocol parsing, and reverse engineering related to Lanzou Cloud within this project were researched and written from scratch independently by the author. No third-party open-source projects or existing implementations were referenced, borrowed, or relied upon. All interface logic is strictly up to date. Therefore, no acknowledgements to other non-fundamental library projects are necessary or made here.
Conclusion
We sincerely welcome anyone to submit Issues to report bugs or open PRs to improve the code. We highly encourage the distribution, packaging, and spreading of this application to help build an open resource-sharing community.
If you find this project helpful, please consider leaving a Star🌟.
설치
This server does not publish a one-line install command.
Open the repository installation guide