ClawHub
Back to skill

Security audit

Docker核心操作指南

Security checks for vulnerabilities and agentic risk

Overview

This is mostly a Docker command guide, but it grants execution-oriented authority while its trigger text is partly unrelated and several destructive Docker operations are under-scoped.

Install only if you want an agent to help with Docker administration. Restrict use to Docker tasks, and require explicit confirmation before commands that delete containers, images, networks, volumes, or run prune/down -v operations, especially on shared or production machines.

SkillSpector

Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • 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
Findings (4)

Description-Behavior Mismatch

Medium
Confidence
95% confidence
Finding
The manifest description mixes Docker administration with unrelated data-analysis/reporting use cases, which can cause the agent to invoke this skill in contexts far outside its stated purpose. In an MD+EXEC skill, scope confusion is dangerous because it may lead to execution-capable Docker guidance being offered during unrelated workflows, increasing the chance of unintended command execution or user misdirection.

Vague Triggers

High
Confidence
97% confidence
Finding
The description contains an overly broad invocation phrase that appears to activate the skill for unrelated tasks such as data analysis and reporting. Because the skill exposes exec/write capabilities and includes many operational Docker commands, overbroad triggering can route unrelated user requests into a powerful command-oriented skill, creating a meaningful risk of inappropriate actions being suggested or executed.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The command section presents destructive cleanup operations like `docker system prune -a` and `--volumes` without an immediate warning adjacent to the commands. In a skill intended for agent use, terse destructive examples can be copied or auto-suggested without the user appreciating that images, containers, networks, and volumes may be irreversibly removed.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The deletion commands for containers, images, networks, and volumes are shown as normal operational examples without clear safety guardrails in the same section. In context, this skill is an execution-capable Docker guide, so presenting destructive commands without warnings increases the likelihood of accidental data loss or service disruption if an agent or user follows them uncritically.

Static analysis

No suspicious patterns detected.