Show HN: MCPSafe – Free security scanner for MCP servers using 5-LLM consensus (mcpsafe.io)

A new tool called MCPSafe has been introduced as a free security scanner for MCP (Multi-Channel Processing) servers, leveraging a unique 5-LLM (Large Language Model) consensus mechanism. Designed for developers preparing to install MCP servers and for registry operators managing safe package catalogs, MCPSafe delivers rapid security assessments. Users can receive a quick verdict in just three minutes, while a comprehensive LLM-based consensus takes about 20 minutes. Public packages are scanned at no cost, and private repositories are available for teams managing their own deployments. This tool is significant for the AI/ML community as it highlights the growing importance of security in AI-driven applications and environments. By scanning over 573 packages and identifying 9,097 vulnerabilities, MCPSafe represents a pivotal step towards enhancing the safety of software installations in the AI ecosystem. The capacity for already-scanned servers to return results in under a second, along with ongoing verification of new commits, ensures that developers have access to up-to-date security information, thus mitigating risks associated with deploying MCP servers.