Scan skills and MCP servers before you install. Detect malware, credential stealers, and data exfiltration in seconds.
Multiple layers of protection to keep your AI agents safe from malicious code.
40+ threat signatures detect environment variable access, data exfiltration attempts, code obfuscation, and known attack patterns—instantly.
Specialized analysis for Model Context Protocol servers. Detect excessive permissions and dangerous tool configurations.
Deep learning models explain what each skill does in plain English, helping you understand exactly what you're installing before you run it.
Three simple steps to secure your AI agent workflow.
Drop a GitHub URL, npm package, or install command. We support multiple input formats.
Our scanner analyzes the code in milliseconds, checking against 40+ threat patterns.
Get a detailed report and verified badge. Install with confidence or skip if flagged.
Instant regex and pattern-based detection. Catches 90% of threats in milliseconds.
LLM-powered analysis that understands code intent and catches sophisticated attacks.
Join developers who trust Vetted.sh to keep their agents safe.