MCP Server Fuzzer
What it is
A black-box security assessment tool for Model Context Protocol (MCP) servers that drives live targets over stdio, HTTP, or SSE transports. It sends realistic and malformed inputs to identify vulnerabilities like tool poisoning, injection, and crashes, then classifies responses and writes detailed findings to disk.
The problem it solves
Lack of specialized tools for black-box security testing of Model Context Protocol servers, which are increasingly used to expose sensitive capabilities to LLMs.
What you could use it for
- Security auditing of custom MCP servers before deployment
- Testing MCP servers for prompt injection and tool poisoning vulnerabilities
- Assessing the robustness of MCP servers against malformed or out-of-contract inputs
- Verifying OAuth boundary security and metadata safety in MCP implementations
Read the source ↗Back to the library →
Summary written by an AI model from the source; check the source before you rely on it.