Skip to main content

Welcome to Helmet Documentation

Helmet is a comprehensive security and management platform for AI infrastructure. It provides visibility, control, and protection for AI-powered applications and their interactions with external tools and resources.

What is Helmet?​

Helmet acts as a security layer and management console for MCP servers, offering:

  • Registry Management: Catalog, scan, and publish MCP servers with built-in security analysis
  • Security Scanning: Automated vulnerability detection, secret exposure prevention, and dependency validation
  • Runtime Monitoring: Real-time logging and monitoring of deployed MCP servers
  • Policy Enforcement: Granular control over tool access and usage patterns
  • Deployment Automation: Streamlined workflows for deploying MCP servers across different environments

Who Should Use Helmet?​

  • Developers building AI applications with MCP servers
  • Security Teams requiring visibility into AI tool usage and potential vulnerabilities
  • Infrastructure Teams managing deployments and runtime environments
  • CISOs and Leadership needing governance and compliance oversight

Key Features​

Security First Approach​

Automated scanning for vulnerabilities, exposed secrets, and unsafe dependencies before deployment.

Centralized Registry​

A unified hub for discovering, managing, and deploying secure MCP servers across your organization.

Real-time Monitoring​

Track every interaction between AI models and your tools with comprehensive logging and analytics.

Policy Management​

Define and enforce rules about which tools can be accessed, by whom, and under what conditions.

Getting Started​

  1. Install the Helmet CLI on your local machine
  2. Configure and run the proxy to start monitoring MCP interactions
  3. Add your first MCP server to the registry
  4. Deploy and monitor your MCP servers

Architecture Overview​

Helmet operates in multiple modes:

  • Proxy Mode: Intercepts and logs MCP protocol communications between clients and servers
  • Agent Mode: Manages deployed servers and enforces policies (coming soon)
  • CLI Mode: Command-line interface for registry, configuration and management (coming soon)

Ready to secure your AI infrastructure? Let's get started with the installation.