ClawHub
Back to skill

Security audit

multi-source-data-cleaner-pro

Security checks for vulnerabilities and agentic risk

Overview

This appears to be a legitimate local data-cleaning skill, but it can save unmasked sensitive sample data in its audit/profile outputs despite advertising PII masking.

Install only if you are comfortable with this skill reading the datasets you explicitly point it at and writing cleaned files plus audit reports. For sensitive customer, employee, medical, financial, or regulated data, use a protected output directory, avoid pii-policy keep unless authorized, and treat profile/audit files as sensitive because they may contain raw sample values.

SkillSpector

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
Findings (3)

Lp3

Medium
Category
MCP Least Privilege
Confidence
89% confidence
Finding
The skill clearly performs local file reads and writes across user-provided datasets and generates multiple output artifacts, yet no explicit permissions are declared. This creates a trust and containment gap: the runtime may grant broader filesystem access than users or policy expect, increasing the chance of unintended access to unrelated local files or overwriting data.

Vague Triggers

Medium
Confidence
78% confidence
Finding
The trigger keywords are broad enough to match many ordinary data-related requests, which can cause the skill to activate in situations the user did not intend. Because this skill reads local files, transforms data, masks or drops fields, and writes outputs, overly permissive invocation increases the risk of unnecessary handling of sensitive datasets or unintended destructive-seeming transformations.

Natural-Language Policy Violations

Medium
Confidence
90% confidence
Finding
The PII hints hard-code Chinese-specific identifiers and address/name assumptions without declaring that detection is China-scoped or providing locale-aware alternatives. In a multi-source data cleaning skill, this can cause systematic misclassification of PII across other regions, leading to missed sensitive data, improper redaction, and compliance gaps when users rely on the detector as a privacy safeguard.

Static analysis

No suspicious patterns detected.