What does the MCP config validator check?
This first MCP Toolkit scanner performs static analysis
of common Model Context Protocol client and server
configuration patterns.
- JSON syntax and configuration shape
- stdio commands and command arguments
- Streamable HTTP and remote URLs
- hard-coded tokens and credentials
- private keys embedded in configuration
- shell execution patterns
- dangerous command-line flags
- private and localhost endpoints
- legacy SSE transport indicators
MCP transports
Local MCP integrations commonly launch servers using
stdio. Remote deployments generally use
Streamable HTTP. Older HTTP+SSE
configurations may still exist for compatibility.
This validator identifies transport indicators but does
not contact remote endpoints during static analysis.
Why scan MCP configuration?
An MCP configuration can grant an AI application access
to local commands, files, APIs, credentials and remote
services. A configuration that looks small can therefore
carry meaningful security impact.
Review the server package, its tool permissions and the
commands it can execute before trusting it with sensitive
systems.
Static analysis limitations
This tool analyzes only the JSON you submit. It does not
prove package integrity, inspect the source code of an MCP
server, verify OAuth behavior or test a remote server for
runtime vulnerabilities.
Those checks will be provided by separate MCP Toolkit
security tools.