ClawHub
Back to skill

Security audit

Cybersecurity Engine

Security checks for vulnerabilities and agentic risk

Overview

This is a markdown-only cybersecurity guidance skill with broad but disclosed security-audit and hardening content, and I found no hidden execution, persistence, credential handling, or exfiltration behavior.

Install only when you want the agent to provide cybersecurity review and hardening guidance. For penetration-testing use, ensure you own the target or have explicit authorization, and define scope before asking the agent to apply the methodology.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

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 (3)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The README presents generic prompts like "Audit security of my Node.js API" and "Harden my AWS infrastructure" as invocation examples, but it does not define when the skill should activate versus when such requests should be treated as ordinary conversation. It also provides no negative examples or explicit trigger constraints, increasing the risk of unintended invocation in normal security discussions.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill metadata description is very broad and framed as generally applicable to security audits, penetration testing, incident response, and building security programs for any stack. In systems that route or auto-select skills from natural language, such expansive wording can cause unintended activation, leading this powerful security-oriented skill to be invoked outside the user's actual intent or appropriate authorization context.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The natural-language command examples are short, generic triggers like 'Audit security of [project/repo]' and 'Pentest methodology for [target]' without guardrails, scope checks, or consent requirements. In agent ecosystems, these phrases increase the chance of accidental or overbroad skill invocation, and because this skill contains high-impact security guidance, misrouting could enable sensitive analysis or offensive-adjacent workflows when not intended.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.