Skillget

superbox.ai

mcp
Guvenlik Denetimi
Uyari
Health Uyari
  • License — License: MIT
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Low visibility — Only 8 GitHub stars
Code Gecti
  • Code scan — Scanned 12 files during light audit, no dangerous patterns found
Permissions Gecti
  • Permissions — No dangerous permissions requested
Purpose
This project provides a centralized registry and command-line interface to discover, deploy, and test Model Context Protocol (MCP) servers. It executes these servers in isolated Cloudflare sandbox environments rather than running them directly on your local machine.

Security Assessment
The overall risk is rated as Low. A code scan of 12 files found no hardcoded secrets or dangerous patterns, and the tool does not request excessive local permissions. By design, the architecture deliberately routes network requests to external Cloudflare Workers and utilizes sandboxed execution (Cloudflare Durable Objects) to prevent untrusted code from impacting your local system. While the CLI interacts with external services and configures AI clients locally, it does so using standard, transparent operations without exposing sensitive data.

Quality Assessment
The project is actively maintained, with its most recent updates pushed today. It is backed by a standard MIT license and includes comprehensive documentation and a clear description. However, community trust and visibility are currently very low. With only 8 GitHub stars, the tool is in its early stages, meaning it has not yet been widely battle-tested by a large user base.

Verdict
Safe to use, though you should expect early-project maturity given its low community adoption.
SUMMARY

📦 Discover, deploy, and test MCPs in isolated sandboxes

README.md 
                          _                      _
                         | |                    (_)
___ _   _ _ __   ___ _ __| |__   _____  __  __ _ _
/ __| | | | '_ \ / _ \ '__| '_ \ / _ \ \/ / / _` | |
\__ \ |_| | |_) |  __/ |  | |_) | (_) >  < | (_| | |
|___/\__,_| .__/ \___|_|  |_.__/ \___/_/\_(_)__,_|_|
         | |
         |_|

CI
PyPI version
PyPI downloads
Python
Go
License

🧰 SuperBox

SuperBox (inspired by Docker Hub) helps you discover, deploy, and test MCPs in isolated sandboxes ( Demo Video ). It includes:

  • A Python (Click) CLI to initialize metadata, run security scans, push to a registry (R2), search, and configure popular AI clients (VS Code, Cursor, Windsurf, Claude, ChatGPT)
  • A Golang (Gin) backend to list/get/create MCP servers with optional pricing and security reports
  • A Cloudflare Worker + Durable Object executor that runs MCP servers on demand directly from their Git repositories using a lightweight TypeScript interpreter (Cloudflare Workers blocks eval() and exceeds the WASM bundle size limit, making Pyodide unusable)

Why this project:

  • There's no centralized MCP registry to discover all MCPs, and many lack clear usage docs.
  • MCPs on our platform pass a 5-step security/quality check (SonarQube, Bandit, GitGuardian) to reduce vulnerabilities and promote best practices.
  • Unlike MCPs that run locally on your machine, MCP servers here execute in sandboxed environments and return responses securely.

Key Features

  • Central MCP Registry: R2-backed registry with per-server JSON for easy discovery and portability.
  • Sandboxed Execution: MCP servers run in Cloudflare Durable Objects and return responses securely. The executor supports requests-based HTTP tools; see cloudflare/README.md for the full scope.
  • Security Pipeline (5-step): SonarQube, Bandit, and GitGuardian checks with a unified report.
  • One-Command Publish: superbox push scans, discovers tools, and uploads a unified record to R2.
  • Client Auto-Config: superbox pull --client cursor|vscode|... writes correct MCP config pointing to the Cloudflare Worker.
  • Terminal Runner: superbox run --name <server> starts an interactive prompt against the Cloudflare executor.
  • Live Logs: superbox logs --name <server> shows instructions for streaming logs via wrangler tail.
  • Tool Discovery: Regex-based discovery across Python code and optional Node package.json definitions.

📚 Documentation

For complete documentation, setup guides, API references, and CLI usage:

🔗 https://superbox.1mindlabs.org/docs

📄 Research Paper

The IEEE research paper for SuperBox is available in the ieee/ directory:

🗂️ Project Structure

.
├── docs/                       # Documentation (INSTALL.md, SETUP.md)
├── ieee/                       # IEEE research paper (paper.pdf, paper.tex)
├── src/
│   └── superbox/
│       ├── cli/                # CLI: init, auth, push, pull, run, search, inspect, test, logs
│       │   ├── commands/       # CLI subcommands
│       │   └── scanners/       # SonarCloud, Bandit, ggshield, tool-discovery
│       ├── server/             # Golang (Gin) app + handlers
│       │   ├── handlers/       # servers, payment, auth, health
│       │   ├── models/         # Request/response types
│       │   ├── helpers/        # Python R2 helper
│       │   └── templates/      # Landing page
│       └── shared/             # Config, models, R2/S3-compat utils
├── pyproject.toml              # Project metadata & dependencies
├── Dockerfile                  # Server container
├── docker-compose.yaml         # Optional local stack
└── tests/                      # pytest suite - see tests/README.md

🌐 API Reference

The HTTP API provides endpoints for server management, authentication, and payments.

For complete API documentation, see:
https://superbox.1mindlabs.org/docs/api

🔧 CLI Overview

The SuperBox CLI provides commands for authentication, server management, and testing:

Authentication:

  • superbox auth register – Register a new account
  • superbox auth login – Log in (email/Google/GitHub)
  • superbox auth logout – Log out
  • superbox auth status – Check authentication status
  • superbox auth refresh – Refresh authentication token

Server Management:

  • superbox init – Initialize a new MCP server project
  • superbox push – Publish server to registry
  • superbox pull – Download and configure server for AI clients
  • superbox search – Search for servers in registry
  • superbox inspect – View server details and security report
  • superbox test – Test server directly from repository (without registry)

Execution & Monitoring:

  • superbox run – Run server in interactive mode
  • superbox logs – View server execution logs

For detailed CLI documentation and usage examples, see:
https://superbox.1mindlabs.org/docs/cli

📦 Installation

pip install superbox

See docs/INSTALL.md for complete installation instructions.

📄 License

This project is licensed under the MIT License.

👥 Authors

Core Contributors:

Acknowledgments:

Yorumlar (0)

Sonuc bulunamadi