Blockbench MCP
About
Blockbench MCP is a plugin and Model Context Protocol (MCP) server that connects AI assistants (like Claude, Cline, or Ollama) to the Blockbench 3D modeling tool. It is designed for Blockbench users who want to control, query, or automate modeling tasks through AI agents.
Details
- Author
- jasonjgardner
- GitHub stars
- 257
- Downloads
- 1,445
- Categories
- Developer Tools
Jump to
- Connects AI assistants to Blockbench via MCP protocol
- Supports multiple MCP clients: Claude, VS Code, Cline, Ollama, and more
- Enables AI-driven 3D modeling commands and queries
- Uses HTTP transport for communication
- Includes a sample project with prompt examples
- Works with Agent Skills to orchestrate tool usage
Setting up with Highlight
This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:
- Download and install Highlight from highlightai.com/download
- Navigate to the plugins tab and select "Add Custom Plugin"
-
Configure the plugin with the settings below
Plugin Name
Blockbench MCPCommand (node, npx, python, etc.)Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.
- Enable "Start Automatically" if you want the plugin to start when Highlight launches
From the repository
Install the Blockbench plugin by opening the desktop version of Blockbench, going to File > Plugins, clicking “Load Plugin from URL”, and pasting the plugin URL. Then configure the MCP server under Blockbench settings (Settings > General) by setting the MCP Server Port and MCP Server Endpoint (default: :3000/bb-mcp). Finally, connect your preferred MCP client (e.g., Claude Desktop, VS Code, Cline, Ollama) to http://localhost:3000/bb-mcp using the provided configuration examples.
Claude Desktop / Cursor
Paste into your MCP client config file to install this server.
{
"mcpServers": {
"blockbench mcp": {
"blockbench-mcp-plugin": {
"command": "npx",
"args": [
"mcp-add",
"--type",
"http",
"--url",
"http://localhost:3000/bb-mcp",
"--scope",
"project"
]
}
}
}
}McpServers
{
"blockbench-mcp-plugin": {
"command": "npx",
"args": [
"mcp-add",
"--type",
"http",
"--url",
"http://localhost:3000/bb-mcp",
"--scope",
"project"
]
}
}Blockbench MCP
https://github.com/user-attachments/assets/ab1b7e63-b6f0-4d5b-85ab-79d328de31db
Plugin Installation
Open the desktop version of Blockbench, go to File > Plugins and click the "Load Plugin from URL" and paste in this URL:
https://jasonjgardner.github.io/blockbench-mcp-plugin/mcp.js
Model Context Protocol Server
Configure the MCP server under Blockbench settings: Settings > General > MCP Server Port and MCP Server Endpoint
The following examples use the default values of :3000/bb-mcp
Installation
General
npx mcp-add --type http --url "http://localhost:3000/bb-mcp" --scope project
VS Code
.vscode/mcp.json
{
"servers": {
"blockbench": {
"url": "http://localhost:3000/bb-mcp",
"type": "http"
}
}
}
Claude Desktop
claude_desktop_config.json (macOS/Linux)
{
"mcpServers": {
"blockbench": {
"command": "npx",
"args": ["-y", "mcp-remote", "http://localhost:3000/bb-mcp"]
}
}
}
claude_desktop_config.json (Windows)
{
"mcpServers": {
"blockbench": {
"command": "cmd",
"args": ["/c", "npx", "-y", "mcp-remote", "http://localhost:3000/bb-mcp"]
}
}
}
Claude Code
claude mcp add blockbench --transport http http://localhost:3000/bb-mcp
Antigravity
{
"mcpServers": {
"blockbench": {
"serverUrl": "http://localhost:3000/bb-mcp"
}
}
}
Cline
cline_mcp_settings.json
{
"mcpServers": {
"blockbench": {
"url": "http://localhost:3000/bb-mcp",
"type": "streamableHttp",
"disabled": false,
"autoApprove": []
}
}
}
Ollama
uvx ollmcp -u http://localhost:3000/bb-mcp
Recommended: jonigl/mcp-client-for-ollama
OpenCode
opencode mcp add
Usage
See sample project for prompt examples.
Skills
Use Agent Skills to orchestrate tool usage.
Plugin Development
See CONTRIBUTING.md for detailed instructions on setting up the development environment and how to add new tools, resources, and prompts.
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
