ClawHub
Back to skill

Security audit

Sql Gen Tool Free

Security checks across malware telemetry and agentic risk

Overview

This SQL helper is mostly coherent, but it asks agents to execute database and shell actions without enough limits or confirmation safeguards.

Review before installing. Use this skill only with least-privilege or read-only database credentials unless you intentionally need writes, and require manual review plus explicit confirmation before running generated SQL against real data. Avoid letting an agent use broad shell access or production credentials automatically.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Context-Inappropriate Capability

Medium
Confidence
90% confidence
Finding
The skill documentation expands a SQL generation tool into command-execution and network-diagnostic behavior, including use of exec and ping-style troubleshooting, without a tight functional need or safety boundary. In an agent context, this broadens the operational surface from text generation into host command execution, increasing the chance of misuse, prompt-driven shell actions, or unintended access beyond SQL authoring.

Missing User Warnings

High
Confidence
96% confidence
Finding
The skill explicitly describes generating SQL for INSERT, UPDATE, and DELETE and elsewhere references execution via exec-capable tooling, yet it does not provide a strong warning or mandatory safeguard for real-database risk. In practice, an agent could generate destructive or privacy-impacting statements and run them against production-like systems, causing data loss, corruption, or unauthorized exposure.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.