ClawHub
Back to skill

Security audit

课程生成器

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed local course-generation workflow that reads user-provided materials, writes course files, and runs a local verifier without network or credential use.

Install only if you want a Chinese-oriented local workflow for converting transcripts or documents into course materials. Point it only at directories you intend it to read, choose an output directory deliberately, review generated manifests for sensitive source details, and run archive or move actions only when you explicitly want files copied or moved into a knowledge base.

SkillSpector

Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
88% confidence
Finding
The skill explicitly instructs the agent to read user-provided files, write course outputs, and execute a local shell verification script, but it does not declare those permissions in a machine-readable way. That creates a transparency and policy-enforcement gap: a host may allow the skill to appear less privileged than it actually is, increasing the risk of unintended file access or command execution.

Natural-Language Policy Violations

Medium
Confidence
93% confidence
Finding
The prompt is explicitly written to produce Chinese-language output and localized formatting by default, but the skill description does not indicate that every user request will always require Chinese. Forcing a language without user opt-in can override user intent, reduce usability, and cause downstream workflow errors when callers expect output in another language.

Static analysis

No suspicious patterns detected.