Security audit

K8s Devops Tool Free

Security checks across malware telemetry and agentic risk

Overview

This Kubernetes manifest skill is mostly purpose-aligned, but it needs review because its write/exec authority is paired with broad activation wording and inconsistent local/API disclosures.

Review this skill before installing in an automated agent. It appears intended for Kubernetes YAML generation, but only use it for explicit Kubernetes manifest tasks, avoid letting it run kubectl apply without direct confirmation, and treat generated Secret manifests and any content sent through the agent LLM as sensitive.

SkillSpector

By NVIDIA

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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 (5)

Description-Behavior Mismatch

Medium

Confidence: 86% confidence
Finding: The documentation is internally inconsistent: it says the free edition does not support custom templates, yet it also advertises import/modify/reset operations and template-related configuration flows. This can mislead an agent or user into enabling functionality beyond the stated trust boundary, increasing the chance of unintended file generation or unsupported execution paths.

Intent-Code Divergence

Medium

Confidence: 88% confidence
Finding: The skill claims to be fully local and API-key free, but also lists an LLM API as a required dependency. This mismatch can cause operators or agents to route sensitive manifest content to external services under the false assumption that processing is local-only, creating confidentiality and trust-boundary risks.

Vague Triggers

Medium

Confidence: 84% confidence
Finding: The activation language is broad enough to match generic coding, debugging, and deployment requests rather than narrowly scoped Kubernetes manifest generation. In agent environments, this can cause over-selection of the skill and unnecessary access to read/write/exec capabilities for prompts that do not require them.

Vague Triggers

Medium

Confidence: 89% confidence
Finding: The capability coverage section enumerates many vague keywords as activation scope, including generic development terms, without meaningful constraints. In a tool-enabled agent, this expands the chance the skill is invoked outside its intended domain, exposing file-writing and command-execution features more often than necessary.

Missing User Warnings

Medium

Confidence: 82% confidence
Finding: The examples describe writing deployment files and output directories but do not warn that existing manifests may be overwritten or that multiple files may be created. In an automated workflow, this omission can lead to accidental destructive changes, confusing repository state, or unsafe application of generated manifests.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.