Security audit

✅ AgentVerif — Scan. Sign. Verify. Control your distribution.

Security checks across malware telemetry and agentic risk

Overview

The skill appears to do the advertised scan, sign, verify, and revoke work, with disclosed API-key and network use, but users should understand that signing changes the supplied ZIP.

Install only if you are comfortable trusting the agentverif-sign Python package and sending skill ZIPs or license IDs to agentverif.com. Do not scan or sign archives containing secrets you cannot share. Keep AGENTVERIF_API_KEY scoped and present only when revoking, and keep an unsigned copy of any ZIP if you need the original preserved.

SkillSpector

By NVIDIA

Vulnerability Patterns

MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (1)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 84% confidence
Finding: The skill advertises access to environment variables and shell-capable behavior but does not declare corresponding permissions. In this context, that matters because the skill explicitly uses an API key for revocation and relies on external tooling, so undeclared env/shell capability weakens transparency and can lead users to authorize or execute higher-risk behavior without clear notice.

VirusTotal

67/67 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.