Introduction:
JFrog Catalog tracks packages used to build Model Context Protocol (MCP) servers, giving security and development teams visibility into a new and rapidly growing AI attack surface
What Are MCP Servers:
Model Context Protocol (MCP) is an open standard that allows AI assistants to connect to external tools and data sources in a standardized way.
An MCP server is a small program that acts as a bridge between an AI assistant and an external system:
[ AI Assistant ] <--> [ MCP Server ] <--> [ Database / Files / Jira /
Artifactory / etc. ]
What Does JFrog Catalog Track?
JFrog Catalog labels packages identified as MCP server building blocks. These are not MCP servers themselves — they are the packages developers use to build MCP servers.
When a package is labeled MCP Servers in Catalog, it means the package is a known SDK, framework, or library used to implement the MCP protocol.
Why MCP Packages Are a Security Concern
MCP servers represent a new attack surface in AI-integrated development environments. A vulnerable or malicious MCP server package can have serious consequences:
-
An AI assistant using a vulnerable MCP server may be manipulated into performing unauthorized actions.
-
Attackers can exploit vulnerable SDKs to exfiltrate data, read secrets, or execute code through AI-initiated requests.
-
MCP servers often run with elevated access to internal systems (databases, filesystems, APIs), amplifying the impact of any compromise.
Example risk chain:
Vulnerable mcp Python package (e.g., v0.9.0 with known CVE)
↓
Developer builds internal MCP server on top of it
↓
AI assistant uses the MCP server to access internal data
↓
Attacker exploits the CVE via AI-triggered requests
How to View MCP Server Packages in Catalog
-
Navigate to Platform > Catalog > Labels>JFrog Labels.
-
Locate the MCP Servers public label card.
-
Click the MCP Server to open the package list.
-
Review packages by type (npm, PyPI, etc.) and version.
-
Click any package to view version details, including scan results.