ClawHub
Back to skill

Security audit

Docker容器专业版

Security checks across malware telemetry and agentic risk

Overview

The skill is broadly consistent with Docker operations, but it warrants Review because it gives agents high-impact Docker, registry, and CI/CD authority with weak scoping and inconsistent credential guidance.

Use this only where you intentionally want an agent to administer Docker infrastructure. Before installing, require explicit host, registry, and environment selection for any deploy, join, push, scale, or CI/CD action, and keep secrets in environment variables or a secret manager rather than local config files.

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)

Intent-Code Divergence

Medium
Confidence
98% confidence
Finding
The skill's API key section says credentials may be stored in local configuration files, while the security section says keys must not be hardcoded into code or configuration files. That contradiction can lead operators or downstream agents to persist registry or webhook secrets insecurely, increasing the chance of credential disclosure from repos, backups, or local file compromise.

Vague Triggers

Medium
Confidence
92% confidence
Finding
The activation text uses broad, repetitive 'Use when' conditions covering security checks, monitoring, logging, alerts, deployment management, and automation without clear boundaries. In an agent setting, this can cause the skill to trigger in overly broad contexts and perform privileged Docker, file, or exec actions when the user did not intend to invoke this capability.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The skill includes commands for cluster initialization, joining nodes, deploying services, pushing images, copying configuration files, and webhook-enabled integrations, but it does not place prominent up-front warnings or approval gates before these system-changing operations. In an agent environment with exec/write tools, that makes accidental or premature infrastructure changes more likely, including production deployment, registry modification, and external notification triggers.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
cicd:
    enabled: true
    pipeline_dir: "./pipelines"
    auto_deploy: false            # 自动部署（谨慎开启）
    environments: ["dev", "staging", "production"]
# ...
  scaling:
Confidence
91% confidence
Finding
The presence of an auto_deploy capability in a CI/CD configuration introduces autonomous state-changing behavior that could deploy builds without sufficient human review if enabled. In a Docker operations skill that already supports exec, write, and deployment flows, this increases the risk of unintended production changes, propagation of vulnerable images, or misuse after prompt-trigger confusion.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.