Elegant Config Guardian

• Outcome: Review
• Latest audit: May 28, 2026, 10:34 AM
• Version:
  0.1.0

Overview

This skill is a plausible config rollback helper, but its main script can run any supplied shell command under the banner of a safety wrapper.

Install only if you fully trust every --apply-cmd passed to this script. Do not build that command from untrusted text, and do not treat the rollback as complete protection because arbitrary command side effects outside the config file are not reverted.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Behavioral AST: exec() Call, eval() Call, Dynamic Import
• 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)

Context-Inappropriate Capability

High

• Confidence: 98% confidence
• Finding: The script executes caller-controlled input through `eval "$APPLY_CMD"`, which allows arbitrary shell execution rather than narrowly applying a configuration change. In a skill advertised as a safe config-rollback helper, this broad execution primitive is especially dangerous because any consumer expecting a constrained config edit path could unintentionally grant full command execution.

Missing User Warnings

High

• Confidence: 96% confidence
• Finding: Using `eval` on a user-supplied argument creates a direct command-injection sink, and the lack of warning or confirmation does not reduce that risk. In this skill context, the script presents itself as a safety mechanism for risky config changes, which makes hidden arbitrary execution more dangerous because operators may trust it for privileged or production use.

VirusTotal

64/64 vendors flagged this skill as clean.