Back to skill

Security audit

Kubernetes Skills

Security checks for vulnerabilities and agentic risk

Overview

This Kubernetes browser automation skill is mostly coherent, but it includes live ArgoCD sync actions and credential examples without enough guardrails.

Review this skill before installing if it may be used against production clusters. Use least-privilege, short-lived credentials, avoid placing real tokens or passwords in prompts or screenshots, and require explicit approval before any ArgoCD sync, deploy, or other cluster-changing UI action.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill demonstrates sending a bearer token in browser headers without warning about secret handling, token scope, or storage risks. In Kubernetes and observability environments, such tokens may grant broad access, and example code can encourage unsafe copying into logs, prompts, screenshots, or source files.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill is presented as general browser automation for Kubernetes web UIs, but it includes an ArgoCD workflow that performs a state-changing sync operation. In an automation context, undocumented or insufficiently signposted mutation capability increases the risk of unintended production changes, especially if an agent selects this skill for read-only inspection tasks.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The ArgoCD login example uses plaintext username and password fields with no warning about credential sensitivity or secure secret handling. In practice, operators may imitate this pattern with real administrative credentials, exposing them to prompt history, screenshots, logs, or accidental reuse in unsafe environments.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The example issues a Synchronize action in ArgoCD without any warning that it can apply configuration changes to a cluster. Because this is a browser automation skill for operational tooling, normalizing one-click state changes without confirmation can lead an agent or user to trigger unintended deployments or reconciliations.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.