Veridict is a Model Context Protocol (MCP) server add-on that provides runtime trust monitoring and reliability scoring for AI agents that call MCP tools. By wrapping an MCP server, Veridict transparently logs and classifies tool execution results, provides both instant trust judgments and detailed statistics, and exposes two tools (`veridict_stats` and `veridict_can_i_trust`) that agents or humans can call to assess server reliability. It integrates with the broader Agent Trust Stack for layered static and dynamic trust analysis, and supports local SQLite logging, CLI inspection, and interoperability with MCP Trust Kit static scans.
Visit Veridict's official website for product details and getting started.