Content
# Tool List
> After installation, you can operate files by chatting with AI, such as reading PDFs, writing Word documents, converting images, unzipping files, and viewing video information, all locally without internet connectivity.
## What is this?
Simply put: **This is a "translator"**.
Usually, to operate files, you need to open various software and click through menus. Now, you just need to tell the AI in the chat box what you want, such as "Convert this image to PNG," and the AI will automatically understand your request and use the appropriate tools to complete the task.
> 🔰 **Completely不懂技术？** No problem. You don't need to understand MCP, stdio, JSON, or other technical terms. Just follow the steps below to install, and then you can talk to the AI.
<details>
<summary><b>🔍 For those who want to know technical details (click to expand)</b></summary>
This is a toolset based on **MCP (Model Context Protocol)**, consisting of 6 independent servers with 25 tools. MCP is an industry-standard protocol for AI, allowing AI models to safely call local programs. All servers in this project follow the MCP stdio specification, and a single deployment can be used in mainstream AI clients such as Cherry Studio, Claude Code, Cursor, and VS Code.
</details>
## What you can do after installation
| You say… | Effect |
|--------|------|
| 🗣️ 「Read contract.pdf」 | AI extracts PDF text |
| 🗣️ 「Check Sheet2 of report.xlsx」 | AI reads Excel data |
| 🗣️ 「Convert photo.png to JPG」 | AI converts image format |
| 🗣️ 「How long is this song?」 | AI displays audio duration and format |
| 🗣️ 「What is the video resolution?」 | AI checks video information |
| 🗣️ 「Unzip archive.rar to D:\output」 | AI unzips file |
| 🗣️ 「Help me write a Word weekly report」 | AI generates .docx file |
| 🗣️ 「Recognize text in this screenshot」 | AI performs OCR text recognition |
> 💡 No need to remember commands or learn syntax. Just talk to the AI like a human. For more examples, see [📖 User Guide](docs/USAGE.md)
## Features
| Category | What can be done | Number of tools |
|------|---------|:--:|
| 📄 **PDF** | Read content, view information (author/pages), generate PDF from Markdown | 3 |
| 📝 **Office** | Read/write Word (.docx), Excel (.xlsx), PowerPoint (.pptx) | 6 |
| 🖼️ **Images** | View information, format conversion, scaling, cropping, OCR text recognition | 4 |
| 📦 **Archives** | List content, unzip (ZIP/RAR/7z/TAR), create archive | 3 |
| 🎵 **Audio/Video** | View audio/video information, format conversion, video screenshot | 4 |
| 🔧 **Script tools** | Execute scripts, view binary files, locally preview web pages, HTML to PDF | 5 |
> A total of **25 tools**, all running locally without internet connectivity. For detailed parameter tables, see [🛠 Tool Reference](docs/TOOLS.md)
## Quick Start (5 minutes)
### Step 1: Install dependencies
```bash
pip install -r requirements.txt
```
### Step 2: Install system tools (as needed)
| If you want to use… | Need to install… | Download |
|-------------|---------|------|
| OCR text recognition | Tesseract | [Download](https://github.com/UB-Mannheim/tesseract/wiki) |
| Audio/video processing | FFmpeg | [Download](https://ffmpeg.org/download.html) |
| RAR/7z unzip | 7-Zip | [Download](https://www.7-zip.org/) |
> If you only need to work with PDFs, Office documents, images, and basic compression, you don't need to install these.
### Step 3: Configure
Open Cherry Studio → **Settings → MCP → Add server**, and fill in the 6 servers according to `cherry_mcp_config.json` (remember to change the paths to your own).
> For detailed installation instructions (including macOS/Linux), see [📥 Installation Guide](docs/INSTALL.md)
### Step 4: Verify
Restart Cherry Studio and enter:
```
What MCP tools can you see?
```
If the AI replies with a list of tools (pdf_read, word_read, image_info…), it means the installation was successful ✅
## Skills
In the "Skills" panel of Cherry Studio, add the following skills to trigger common operations with one click:
| Skill name | Trigger word | Effect |
|--------|--------|------|
| Read PDF | `Read PDF` | Read and summarize PDF content |
| PDF to text | `PDF to text` | Extract full PDF text |
| Image information | `Image information` | View image dimensions and format |
| Batch format conversion | `Batch format conversion` | Convert image formats in a folder |
| Unzip file | `Unzip file` | Unzip archive to a specified directory |
| Video information | `Video information` | View video resolution and duration |
## Usage in other AI tools
In addition to Cherry Studio, this project can also be integrated with:
**Claude Code · Cursor · Continue · Windsurf · Cline · VS Code / Copilot · Claude Desktop**
Each client only requires a few lines of JSON configuration. For details, see [🔌 Multi-platform configuration](docs/INTEGRATIONS.md)
## Project structure
```
cherry-studio-mcp/
├── README.md ← You are reading this
├── requirements.txt ← Python dependency list
├── cherry_mcp_config.json ← Configuration template (paths need to be modified)
├── servers/
│ ├── pdf_server/server.py ← PDF processing
│ ├── office_server/server.py ← Office document processing
│ ├── image_server/server.py ← Image processing
│ ├── archive_server/server.py← Archive processing
│ ├── media_server/server.py ← Audio/video processing
│ └── script_server/server.py ← Script/tool processing
└── docs/
├── USAGE.md ← 📖 User Guide (recommended to read first)
├── INSTALL.md ← 📥 Installation and configuration
├── INTEGRATIONS.md ← 🔌 Multi-platform integration
└── TOOLS.md ← 🛠 Tool parameter reference
```
## Legal statement and open-source license
### 📜 Open-source license
This project uses the **Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0)** license.
| You can… | You cannot… |
|----------|-----------|
| ✅ Free download and use | 🚫 Used for commercial purposes |
| ✅ Free modification of source code | 🚫 Integrated into commercial products for sale |
| ✅ Free sharing with others | 🚫 Providing paid services based on this project |
| ✅ Used for learning, research, and education | 🚫 Any use for monetary compensation |
> 📄 The full legal text is in the [LICENSE](LICENSE) file. The Chinese summary is in the [NOTICE](NOTICE) file.
### 🛡️ Legal protection statement
```
╔══════════════════════════════════════════════════════════════╗
║ ⚠️ This project is protected by international copyright law and CC BY-NC 4.0 license ║
║ ║
║ ● Copyright holders reserve the right to pursue legal action against infringement ║
║ ● Strictly prohibited for any form of commercial use, violators will be held accountable ║
║ ● Commercial authorization requires separate contact with copyright holders for written permission ║
║ ● This project is provided "as is," without any form of warranty ║
║ ● Users must ensure they have legitimate rights when using this project to process files ║
╚══════════════════════════════════════════════════════════════╝
```
### 📋 Third-party dependency statement
This project uses the following open-source libraries, each with its original license:
- PyMuPDF (AGPL/Commercial), python-docx (MIT), openpyxl (MIT)
- python-pptx (MIT), Pillow (HPND), pytesseract (Apache 2.0)
- ReportLab (BSD), FFmpeg (LGPL/GPL), Tesseract (Apache 2.0)
> Users must comply with the license terms of the above third-party libraries.
### 📞 Contact information
- 🐛 Technical issues: [GitHub Issues](https://github.com/ChinaPxs/cherry-studio-mcp/issues)
- ⚖️ Legal/commercial authorization: Please submit through GitHub Issues with the title "Commercial authorization inquiry"
