ClawHub
Back to skill

Security audit

cdn

Security checks for vulnerabilities and agentic risk

Overview

This CDN skill is not clearly malicious, but it asks for broad command/API/file authority without enough scoping or user-control guidance for infrastructure-changing actions.

Review this skill before installing. It appears intended for CDN operations, but only use it where you are comfortable allowing shell commands and CDN-provider API actions, and require explicit confirmation before cache purges, configuration changes, file writes, or authenticated requests.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (4)

Context-Inappropriate Capability

Medium
Confidence
95% confidence
Finding
The skill declares generic `exec` capability even though the documented purpose is CDN configuration and troubleshooting, and the manifest provides no scope limits, allowlist, or approval guardrails for what commands may be run. In an agent setting, this creates an unnecessary path from natural-language CDN requests to arbitrary system command execution, increasing the risk of filesystem access, secret exposure, and host-side actions unrelated to CDN tasks.

Intent-Code Divergence

Medium
Confidence
91% confidence
Finding
The documentation asserts that commands run in a 'safe sandbox', but the manifest does not define any actual sandboxing, isolation boundary, or execution constraints. This can mislead users and downstream agents into trusting risky behavior, lowering caution around commands that may affect the local system, network, or secrets.

Vague Triggers

Medium
Confidence
84% confidence
Finding
The activation guidance is overly broad, including generic coding, debugging, deployment, and any CDN-related need, which can cause the skill to be invoked in contexts beyond its narrow purpose. Because the skill also advertises powerful tools like `exec`, broad triggering raises the chance that high-privilege capabilities are loaded for ambiguous or unrelated requests.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill describes file handling, API integration, and command execution capabilities, including execution in a purported sandbox, but does not clearly warn users that these actions may change system state, contact external services, or expose credentials. In a skill that may be auto-invoked, missing user-facing safety warnings materially increases the risk of unintended destructive actions or secret leakage.

Static analysis

No suspicious patterns detected.