Independent intelligence on the MCP ecosystem
Original analysis derived from exact-version inventory, isolated runtime verification, capability history, and publisher attribution across the MCP ecosystem.
Intelligence reports
Companies With the Most Community-Associated MCP Servers
Compare company product ecosystems by community publication, retained official attribution, distribution presence, and current-version verification coverage.
State of MCP Security 2026
The market, publisher-attribution, remote-exposure, observed-tool, vulnerability, and verification findings that define the current MCP security landscape.
MCP Security Scanner Coverage: What We Tested and Proved
Exact counts for npm, PyPI, and remote coverage; current-version outcomes; retained vulnerability techniques; methodology; and explicit evidence limits.
MCP Security Landscape: vulnerabilities, coverage, and risk
What current exact-version verification proves, what remains unknown, and where MCP security evidence is most consequential.
MCP Ecosystem Threat Map
Decision-grade analysis of identity fragmentation, remote exposure, and high-authority capability concentration derived from the live catalog relationship graph.
Cross-repository MCP risk is the missing security boundary
A safe-looking server in one repository can become the read step in an attack path completed by a second server elsewhere. Repo-by-repo controls cannot see the composition.
Unofficial MCPs create a brand security perimeter companies do not control
Our catalog links brand-associated names to publisher identity, exact versions, observed tools, and reproduced findings—revealing customer risk outside the vendor’s repositories.
What our runtime verification data reveals about MCP security
Live coverage, proven findings, and the gap between source claims and behavior observed while executing exact releases.