ClawHub
Back to skill

Security audit

Engineering Discipline

Security checks for vulnerabilities and agentic risk

Overview

The reported skill appears to be prompt guidance only, with no evidence of malware, credential access, destructive behavior, or hidden execution.

This looks acceptable to install as a prompt-only coding guidance skill. Review its activation wording before installing, especially if you work in multilingual contexts, because broad triggers may influence the agent more often than intended.

SkillSpector

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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)

Vague Triggers

Medium
Confidence
93% confidence
Finding
The skill is designed as a permanent behavior modifier and its description embeds many broad trigger phrases such as generic coding and production-quality terms. That makes accidental invocation plausible during normal engineering conversations, which can unpredictably alter agent behavior and create prompt-scope confusion even if the guidance itself is not overtly malicious.

Natural-Language Policy Violations

Low
Confidence
80% confidence
Finding
The description mixes English and Chinese trigger phrases without defining language scoping, precedence, or opt-in behavior. In multilingual environments this increases the chance of unintended activation from ordinary discussion, translation, or copied text, expanding the prompt-injection surface and making behavior less predictable.

Static analysis

No suspicious patterns detected.