ClawHub
Back to skill

Security audit

Skill Org

Security checks for vulnerabilities and agentic risk

Overview

This is a markdown-only organization-design skill that helps draft job descriptions and skill files, with no executable code or hidden access requests.

Installers should be aware that this skill may activate for broad planning or task-assignment requests; review generated JD or skill text before relying on it, especially when defining authority boundaries for future agents.

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 (2)

Vague Triggers

Medium
Confidence
95% confidence
Finding
The skill description embeds very broad trigger phrases such as '帮我写一个skill', '拆分这个工作流', and '这个任务该谁负责', which are common in ordinary collaboration and planning conversations. This can cause unintended invocation of the skill, routing user requests into an organizational-design workflow when the user may have intended a narrower or different task, increasing mis-execution risk and making downstream agent behavior less predictable.

Vague Triggers

Medium
Confidence
97% confidence
Finding
The instruction that the description must cover both exact and 'fuzzy' trigger phrases explicitly encourages broad matching on ambiguous user language. In a dispatcher or agent-routing system, this weakens boundary clarity and raises the chance that unrelated requests will activate this skill, potentially overriding more appropriate specialized skills or leading to incorrect workflow compilation.

Static analysis

No suspicious patterns detected.