Qiniu MCP Server

What is Qiniu MCP Server?

Qiniu MCP Server is a Model Context Protocol (MCP) server built on Qiniu Cloud products, enabling AI large language model (LLM) clients to access Qiniu Cloud Storage, intelligent multimedia services, and CDN capabilities within the AI context. It runs as a Python-based MCP server and works with MCP-compatible AI clients like Cline, Cursor, and Claude.

How to use Qiniu MCP Server?

Install Python 3.12+ and the uv package manager, then configure environment variables (QINIU_ACCESS_KEY, QINIU_SECRET_KEY, QINIU_REGION_NAME, QINIU_ENDPOINT_URL, and QINIU_BUCKETS). In an MCP client such as Cline (VS Code extension), add a server entry pointing to uvx qiniu-mcp-server, set the environment variables, and enable the MCP server. Start chatting with the AI and use natural language commands like "list buckets", "read file content", or "resize an image".

Key features of Qiniu MCP Server

• Retrieve bucket and file lists from Qiniu Cloud Storage
• Upload local files or file content to buckets
• Read file content and generate download URLs
• Resize images and apply rounded corners
• CDN file refresh and prefetch by URL

Use cases of Qiniu MCP Server

• An AI assistant lists all storage buckets and their files on demand
• A developer uploads a configuration file to a specific bucket via chat
• An operator reads a log file stored in Qiniu without leaving the AI client
• A content manager resizes an image and refreshes the CDN link immediately
• A team pre-fetches critical assets across multiple buckets using natural language

FAQ from Qiniu MCP Server

What credentials are required to use Qiniu MCP Server?

You need Qiniu Access Key and Secret Key, plus the region name and endpoint URL for your storage. These are set via environment variables.

Which AI clients does Qiniu MCP Server support?

It works with any MCP-compatible client, including Cline, Cursor, and Claude. The README provides explicit configuration for Cline and notes that Claude may require the absolute path to uvx.

How many buckets can I configure at once?

You can configure multiple buckets by separating their names with commas in the QINIU_BUCKETS environment variable. It is recommended to configure no more than 20 buckets.

Can I develop or extend the server myself?

Yes. Clone the repository, create a uv virtual environment, install dependencies, and add new tools/resources in the core directory following the pattern shown (e.g., a new storage package). The server also supports SSE transport for web applications.

What transport modes does the server support?

It supports stdio (default) and SSE modes. For SSE, use --transport sse --port 8000 when starting the server.