ClawHub
Back to skill

Security audit

蜂鸟云号管理专家 | 青虎AI

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed Qinghu/Fengniao phone-number management workflow with real account and money actions, but it includes clear confirmation and authorization controls.

Install only if you intend to use Qinghu/Fengniao cloud-number services. Expect to provide or expose a Qinghu API token, and review every confirmation carefully before purchases, payments, recharge, withdrawal, renewals, returns, deletion, activation, or auto-renew changes. Be aware the skill may activate on broad number/SMS-registration wording, so decline tool authorization if it was triggered unintentionally.

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 (1)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill’s trigger is explicitly mandatory and very broad, covering generic phrases about buying phone numbers, receiving SMS codes, store registration, recharge, and withdrawal. That creates a real risk of unintended invocation in unrelated conversations, causing the agent to enter a sensitive financial/account-management workflow and potentially collect or act on high-risk account data when the user did not intend to use this skill.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.