Canopii Trust Index is a security-scored registry of MCP servers. It assigns one score per version and is described as scanning servers before agents connect, with a focus on tool poisoning, prompt injection, supply-chain risk, and credential risk.
The service shows a live ecosystem view rather than a static list. It says all servers are independently scanned and scored, and it publishes counts for scored servers, grade distribution, high-risk servers, live-verified endpoints, and unverifiable servers. The page also presents a browsable list of servers and a section for top-scored entries, along with a methodology page.
API access is part of the product. The page says users can check any MCP server’s security score from their own tools, including for CI gates, procurement review, and agent allow-lists. That API is key-authenticated and rate-limited. The interface shown is web-based, with links to browse servers online and request API access.
The page names an Enterprise section, but it does not provide pricing or licensing terms. Its scope is the MCP ecosystem, and its purpose is to surface security scores for MCP servers so their risk posture can be reviewed before use.
Canopii Trust Index is a LLM eval & observability project. It focuses on assessing and mitigating security risks of MCP servers before agent connection. It is built as a B2B product for AI infrastructure operators and security teams. Canopii Trust Index follows a freemium model, with paid plans starting at $49. It ships for the web and API.
It is developed by Canopii, and it first shipped in 2024. Among its 8 catalogued features are security scoring, continuous scanning, and API access. It exposes integrations via an MCP server and a public API.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do