Arcs MCP

🚀 Arcs-MCP

A multi-platform one-click publishing tool supporting MCP services that simplifies the process of sharing and publishing articles across multiple platforms.

🚀 Quick Start

If you need to share and publish articles to multiple platforms, this tool can simplify the process. It supports MCP services, allowing users to use natural language processing and achieve one-click publishing of articles on multiple platforms.

✨ Features

• Simplify the process of sharing and publishing articles to multiple platforms.
• Support MCP services, enabling natural language processing for one-click publishing.
• Provide a web page for publishing operations, allowing users to upload local .md files.

🔑 Prerequisites

• AI clients compatible with MCP: Claude Desktop, Gemini CLI, Cherry Studio, or other MCP clients.
• uv: A modern Python package installer and resolver.

📦 Installation

1. Install the uv Environment

It is recommended to use the Cherry Studio client to configure the MCP service of this tool, as the operation and environment configuration will be more user-friendly. Cherry Studio also comes with the deployment function of the uv environment, allowing users to complete the installation with one click.

💡 Usage Tip

Currently, Cherry Studio only uses the built-in uv and bun and will not reuse the uv and bun installed in the system.

You can also choose to manually install the uv environment independently and complete the deployment using the command line:

macOS & Linux

# Use `curl` to download the script and execute it via `sh`:
curl -LsSf https://astral.sh/uv/install.sh | sh

# If your system does not have `curl`, you can use `wget`:
wget -qO- https://astral.sh/uv/install.sh | sh

Windows

# Use `irm` to download the script and execute it via `iex`:
powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

2. Clone the Repository

1. Clone the repository and switch to the directory:

git clone https://github.com/Cyanty/Arcs-MCP.git
cd Arcs-MCP

2. Copy the environment configuration file:

cp .env.example .env

💡 Usage Tip

Some platforms support API article publishing. You can configure the required platform keys through the .env file.

3. Build and run the service:

uv run --directory your/path/to/Arcs-MCP server.py

3. Configure the MCP Client

Configure your MCP client. Here, we take the AI client tool of Cherry Studio as an example.

• Click the Quick Create option in the Add Server button, select the streamableHttp type, and add the following URL:

http://127.0.0.1:8001/submit/mcp

The Quick Create configuration is as shown below:

Or, use JSON configuration:

• Click the Import from JSON option in the Add Server button and add the following content to the configuration file:

{
  "mcpServers": {
    "SubmitArticleServer": {
      "type": "streamableHttp",
      "url": "http://localhost:8001/submit/mcp"
    }
  }
}

💻 Usage Examples

Basic Usage

💡 Usage Tip

Before starting, please ensure that your MCP client has been successfully configured as described above.

Available Tools

A list of available tools:

Tool descriptions:

• help_open_browser: Helps the user open the browser (MCP publishing browser) once. For example, "Please help me open the browser." Users can use this browser to perform operations such as account login.
• submit_verify_login: Verifies the account login status of all publishing source platforms. For example, "Please help me verify the login status of all platforms." Users can use this method to verify the login status of each platform account and the validity of personal tokens/keys.
• get_submit_toggle_switch: Gets the current publishing switch status of each publishing source. For example, "Please tell me the current publishing switch status of each platform." The publishing switch is used to indicate to the user which platforms are available for article publishing.
• update_submit_toggle_switch: Updates the publishing switch status of each publishing source platform. For example, "Please turn on the publishing switch for the CSDN platform." Turning on the platform publishing switch can be used to publish articles to the specified platform.
• submit_article_content_prompt: Publishes the text content of an article to each platform. For example, "Publish the text content of this article to CSDN." After waiting for the publishing operation to complete, you can see your article on the publishing platform.
• submit_article_file_to_platforms: Publishes an article to each platform through the file path. For example, "Help me publish the article at /your/path/to/file-absolute-path to CSDN." After waiting for the publishing operation to complete, you can see your article on the publishing platform.

Article Publishing Example

💡 Usage Tip

Before publishing, please ensure that your MCP publishing browser and personal tokens/keys are in a logged-in or available state.

Use Cherry Studio to publish an article to CSDN via chat.

In the new chat window, click on the MCP server and select the previously configured SubmitArticleServer MCP service:

In the chat box, enter, for example: "Help me publish this article to CSDN. The local path of the article is: C:\Users\Administrator\Desktop\发布一篇文章测试.md". Wait for the large model to return the publishing result:

This publishing tool also supports publishing in text content format. However, due to differences in the processing methods of integrated AI client tools, the large model may read the entire content of the article as context, which consumes more Tokens and takes longer. It is recommended to use the file path method to publish articles.

An example of chat input for publishing in text content format is as follows:

🔧 Technical Details

This publishing tool supports MCP services and achieves efficient publishing by integrating the publishing interfaces of various platforms and automation technologies:

• Unified Entry: Supports direct import of content from .md files and adapts to MCP services for asynchronous parallel publishing.
• Image Processing: Automatically converts external link images to platform-compatible image link addresses.
• Publishing Methods: Adopts two strategies, API calls or browser automation, according to the characteristics of different platforms.

Currently supported publishing platforms (other platforms can be horizontally extended customarily)

In addition, the tool also provides a web page for publishing operations, allowing users to perform operations on the browser page and upload local .md format files to publish articles on various platforms.

For example, visit http://127.0.0.1:8001, and the publishing page is as follows:

🤝 Contributing

Contributions are welcome! Whether it's fixing bugs, adding new features, or improving the documentation, you can submit a Pull Request or open an issue at any time.

📄 License

This project is licensed under the Apache License 2.0. See the LICENSE file for the full text.