I See You

Scanning AI marketplaces for malicious packages. Prompt injection, data exfiltration, and obfuscation — detected and exposed.

12Critical threats

19Total findings

30Packages scanned

7Marketplaces

Detection over time (30 days)

Findings by Category

Recent Critical Threats

Risk	Package	Marketplace	Author	Date
critical	ai.smithery/ChiR24-unreal_mcp_server	MCP Registry	smithery	Today
critical	ai.smithery/ChiR24-unreal_mcp	MCP Registry	smithery	Today
critical	blockscout/mcp-server	Smithery	blockscout	Today
critical	vllm	PyPI	vLLM Team	Today
critical	@syke1/mcp-server	npm	khalomsky	Today
critical	@kickflow/mcp-server	npm	GitHub Actions	Today
low	@drawio/mcp	npm	drawio	Today
critical	@hypertrack/mcp-server	npm	hypertrack_devops	Today
critical	fastapi	PyPI	Unknown	Yesterday
critical	modal	PyPI	Unknown	Yesterday

Risk Distribution

Critical

High

Medium

Low

Clean

Marketplace Safety Scores

MCP Registry

SkillsMP

100%

PulseMCP

100%

npm

PyPI

Smithery

Glama

33%

We all see you

We're early and we'll get things wrong. Help us get it right — report false positives, contribute rules, and shape a more precise scanner together.