ClawHub
Back to skill

Security audit

Cloud IAM Auditor / 云权限审计助手

Security checks across malware telemetry and agentic risk

Overview

This IAM audit skill is mostly coherent, but it advertises automated cloud permission changes without clear approval, dry-run, or rollback safeguards.

Before installing, treat this as an audit-first skill. Do not allow it to change IAM policies automatically unless you have reviewed the proposed diff, tested in dry-run mode, limited it to specific accounts or policies, and confirmed rollback access.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill advertises automated remediation that can apply least-privilege IAM policy changes without manual intervention, but it does not warn users that IAM modifications are high-impact and can break services, lock out users, or cause broad operational disruption if recommendations are wrong. In a security/audit skill, presenting automatic policy application as a feature without explicit guardrails increases the risk that operators will trust and execute impactful changes too casually.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.