XLUXX

XLUXX is an AI trust intelligence service for Model Context Protocol servers. It scores, audits, and tracks MCP servers so developers and enterprises can assess which external tools are safe to connect to AI assistants before using them in production.

The service says it operates a trust scoring system for MCP servers and publishes the results in real time through a public API. Its process includes crawling official and community MCP registries every six hours, fingerprinting newly found servers, and extracting declared capabilities, tool descriptions, authentication requirements, and transport protocols. It also runs live behavioral tests to check how servers respond, whether they time out, whether they make unexpected outbound connections, and whether their responses match their documentation. Those findings feed a multi-factor trust score that includes uptime and reliability, known CVEs or vulnerability patterns, PII detection in outputs, behavioral consistency across repeated queries, and registry provenance. XLUXX also runs daily security audits and offers a public leaderboard of the most trusted and most-used servers.

The API supports bulk queries, historical score timelines, and webhook alerts when a server’s trust status changes or a previously trusted server is flagged for a new vulnerability. The page says developers building AI pipelines, enterprises deploying AI agents, and security teams auditing AI usage are the intended users. It also states that the catalog currently covers more than 15,000 MCP servers and over 7,200 individual tools.

Access is provided through a public REST API, and the page advertises 100 free API requests per day for scoring MCP servers. The site also links to API docs, dashboard, compare, blog, newsletter, AI providers, AI guides, AI glossary, and SDK references for npm and Python, along with a GitHub link. XLUXX describes itself as a neutral trust layer for AI systems and compares its role to a certificate authority for AI agents.