MCPSpy
Kernel-level monitoring and security analysis of MCP traffic
All Skills in This Repository (4)
Pure Emerald Level Indicatorsgit-commit-creator
Create conventional Git commits with feature branches and staged changes.
go-testing
Run and write Go tests for MCPSpy with require and assert conventions.
github-issue-creator
Create GitHub issues for MCPSpy with conventional titles and body files.
security-integration-tests
Runs and debugs prompt injection detection integration tests against the HuggingFace Inference API.
Frequently Asked Questions
FAQPage SchemaHow to install MCPSpy?▼
Run `npx skills add alex-ilgayev/MCPSpy --all -g -y` in your terminal to install all skills in this suite globally.
What does MCPSpy monitor?▼
It uses eBPF to capture MCP JSON-RPC messages over stdio and HTTPS transports in real time, showing requests, responses, and tool calls between clients and servers.
Can MCPSpy detect prompt injection attacks?▼
Yes. With the --security flag and a HuggingFace token, it scans tool call arguments and responses for injection and jailbreak attempts, assigning risk levels from low to critical.
Does MCPSpy work in Kubernetes?▼
Yes. It deploys as a DaemonSet to monitor MCP traffic from AI services like LangFlow and LangGraph across your cluster.
What are the system requirements for MCPSpy?▼
You need Linux kernel 5.15 or later and root privileges, since eBPF requires kernel-level access. Pre-built binaries are available for linux-amd64 and linux-arm64.
Related Repositories in Software Engineering
View All in Software Engineering→openclaw
Run a personal AI assistant across your devices and chat apps
superpowers
Gives coding agents a disciplined workflow from idea to merged code
react
AI agent skills for building, testing, and porting React core