Pinme Deploy

You deploy static sites to IPFS with built‑in Pinme support through this MCP server. It lets you upload files or entire build folders, automatically detect common output directories, check deployment status, and review a history of deployed projects.

How to use

You interact with this MCP server from your MCP client to upload content to Pinme and manage deployments. Two common workflows exist: uploading via an MCP command that runs locally, or using a preconfigured MCP server entry that you launch from your client. After a successful upload you receive a preview link, an IPFS CID, and an ENS URL for the deployed project. You can also check status by CID or ENS URL and list all deployments to review past entries.

How to install

Prerequisites you need before starting are Node.js and npm, plus a working internet connection for installing packages.

Option 1. Run directly with npx (recommended, simplest) in your MCP client configuration:

{
  "mcpServers": {
    "pinme-deploy": {
      "command": "npx",
      "args": ["-y", "git+https://github.com/liaoshengrong/pinme-deploy-mcp.git"]
    }
  }
}

Option 2. Globally install the MCP server via npm and reference it from your client configuration

Install the MCP server globally with npm and then configure the client to call the installed binary.

npm install -g github:liaoshengrong/pinme-deploy-mcp

Option 3. Use the installation script

Run the installation script directly to set up the MCP server on your system.

curl -fsSL https://raw.githubusercontent.com/liaoshengrong/pinme-deploy-mcp/main/install.sh | bash

Option 4. Local development installation

If you are developing locally, clone the project, install dependencies, and build the distribution. Then reference the built file by its absolute path in your MCP client.

git clone https://github.com/liaoshengrong/pinme-deploy-mcp.git
cd pinme-deploy-mcp
npm install
npm run build

Note on running as an MCP server

To run as a server in MCP you typically specify the runtime command and its arguments so the client can launch it. For example, you can run the built entry point with Node using an absolute path to the compiled file.

Configuration in your MCP client

You can configure multiple MCP server entries to support different deployment methods. The following examples illustrate possible configurations you may use from your MCP client.

Examples of common MCP server configurations

{
  "mcpServers": {
    "pinme_npx": {
      "type": "stdio",
      "name": "pinme_npx_mcp",
      "command": "npx",
      "args": ["-y", "git+https://github.com/liaoshengrong/pinme-deploy-mcp.git"]
    },
    "pinme_binary": {
      "type": "stdio",
      "name": "pinme_mcp",
      "command": "pinme-deploy-mcp",
      "args": []
    },
    "pinme_local_dist": {
      "type": "stdio",
      "name": "pinme_mcp_local",
      "command": "node",
      "args": ["/absolute/path/to/pinme-deploy-mcp/dist/index.js"]
    }
  }
}

Available tools

deploy_to_pinme

Uploads a file or directory to Pinme. If no path is provided, common build output directories are auto-detected.

check_pinme_status

Checks the status of a deployed project by CID or ENS URL.

list_deployments

Lists all previously uploaded projects with metadata such as CID, ENS URL, and sizes.