ClawHub
Back to skill

Security audit

Alibabacloud Cadt Probe

Security checks across malware telemetry and agentic risk

Overview

This is a legitimate Alibaba Cloud inventory helper, but its permissions and output can reveal sensitive infrastructure details.

Install only for Alibaba Cloud accounts where you intend the agent to inspect infrastructure inventory. Prefer scoped RAM permissions, explicit --regions and --list-types filters, avoid --debug unless troubleshooting, and protect any exported JSON as sensitive infrastructure data.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Context-Inappropriate Capability

Medium
Confidence
91% confidence
Finding
The permission-failure workflow explicitly directs the skill to invoke a separate `ram-permission-diagnose` skill, creating an undeclared cross-skill dependency and expanding the operational scope beyond simple resource probing. This can lead to privilege-escalation-by-proxy or unintended disclosure if the secondary skill has broader capabilities or weaker safeguards than this skill's stated purpose.

Vague Triggers

Medium
Confidence
83% confidence
Finding
The trigger list includes very generic terms such as 'probe', 'list resources', and 'view cloud resources', which can cause the skill to activate for broader requests than intended. In this context, accidental invocation is risky because the skill can enumerate cloud infrastructure and emit sensitive inventory data.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The skill promotes exporting full cloud resource inventory to JSON/files without a clear warning that the output may contain sensitive infrastructure metadata, identifiers, topology, and relationship data. This increases the chance of unintentional disclosure through local files, logs, shared workspaces, or downstream agent processing.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The document enumerates broad read-level permissions that expose full cloud inventory, network topology, security group relationships, whitelists, and related configuration metadata, but it does not prominently warn users about the sensitivity of the resulting data. In a resource-discovery skill, this omission increases the chance that operators grant broad access without understanding the privacy and security exposure of account-wide reconnaissance data.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The --debug option prints API response metadata and body content to stderr, and those responses can contain discovered cloud resource inventory and relationship data. In an agent or automated execution context, stderr is often captured into logs, traces, or orchestration systems, so enabling debug can unintentionally exfiltrate sensitive infrastructure information without an explicit runtime confirmation or redaction layer.

VirusTotal

61/61 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.