ClawHub
Back to skill

Security audit

AuditClaw Azure

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed Azure compliance-auditing helper that uses read-only Azure access and stores results locally for its stated GRC purpose.

Install only if you are comfortable granting read-only subscription-wide Azure visibility to this tool. Use the minimum documented Reader and Security Reader roles, keep the Azure client secret secure, and protect or back up the local GRC database because it can contain infrastructure and security findings.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Lp3

Medium
Category
MCP Least Privilege
Confidence
87% confidence
Finding
The skill declares itself as user-invocable and executable, requires sensitive Azure credential environment variables, and directs use of shell/Python commands, yet it does not declare explicit permissions for those capabilities. This creates a transparency and policy-enforcement gap: users and any host permission model may not be clearly informed that the skill accesses secrets from the environment and executes local commands.

Tp4

High
Category
MCP Tool Poisoning
Confidence
91% confidence
Finding
The skill is described as performing read-only Azure evidence checks, but the documentation also states it writes results and integration status into a shared local SQLite database and performs connection probing. That mismatch can mislead users about side effects and data handling, which is security-relevant because persisted compliance data may contain sensitive infrastructure details and may be accessed by other tools sharing the database.

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The skill repeatedly states that evidence is stored in a shared GRC database at ~/.openclaw/grc/compliance.sqlite, but it does not present this as an explicit user warning or consent point. Because the stored evidence includes JSON check results about Azure security posture, this persistence can expose sensitive configuration data to other local users, processes, backups, or adjacent skills that can read the same database.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal