Data Governance

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:22 PM UTC
• Version:
  1.10.1

Overview

This is a disclosed data-governance helper that connects to user-specified databases to inspect schemas and selected table data, with no evidence of hidden execution, exfiltration, mutation, or persistence.

Install this only for databases you own or are authorized to inspect. Use a dedicated read-only database account scoped to the needed tables, avoid putting real passwords in command-line connection strings, and review generated reports before sharing because they may contain schema names, sensitive field names, or sampled data-derived quality findings.

SkillSpector

By NVIDIA

Vulnerability Patterns

• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (2)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 86% confidence
• Finding: The skill documentation instructs users to set database credentials via environment variables, install packages, and run scripts that connect to databases, but the skill does not declare corresponding permissions. Undeclared access to env, file writes, and network/database connectivity reduces transparency and can lead to unexpected credential use or data access in environments that rely on permission declarations for trust and sandboxing.

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 92% confidence
• Finding: The skill is presented as a data governance and asset-management helper, but its usage instructions show direct database access and possible reading of schema and table data without clearly disclosing that behavior in the description. This mismatch is dangerous because users may invoke the skill expecting advisory/reporting behavior while actually granting it access to sensitive production data, increasing the risk of unintended exposure or over-privileged execution.

VirusTotal

61/61 vendors flagged this skill as clean.