threat.cafe

This open threat-intelligence service checks AI-agent packages, skills, MCP servers, and configuration files against a reviewed, signed blocklist, offering a local pre-run hook, fast offline lookups, an API, evidence-linked records, and customizable policies to block, warn, log, or request approval. It helps prevent malicious tools, fake installers, compromised dependencies, and other agent supply-chain threats without sending private project details over the network, and is intended for developers, security and IT teams, managed service providers, researchers, and agent-platform operators.

threat.cafe kann gefunden werden in Bot Management & Anti-Bot Kategorien.