ClawHub
Back to skill

Security audit

抖音-爆款视频跟卖与铺货专家 | 青虎AI

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed Douyin commerce workflow that uses Qinghu/Xiaofeng APIs, with some normal but sensitive handling of API tokens, store actions, and exported data files.

Before installing, confirm you are comfortable granting Qinghu/Xiaofeng API access for Douyin store workflows, prefer environment or managed secrets over pasting tokens into chat, review any store-distribution preview before execution, and remember that larger datasets may be saved as local files.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Vague Triggers

Medium
Confidence
95% confidence
Finding
The trigger description is extremely broad and built around common e-commerce terms such as '上架', '采集', '抖音', and '抖店', plus a 'must trigger' directive. In an agent environment, this can cause the skill to activate in many adjacent commerce conversations where the user did not specifically request this workflow, increasing the chance of unintended API use, data handling, or operational actions.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill directs the agent to automatically export datasets to local files and reuse cached local file paths without first warning the user that data will be written to the host environment. This creates a risk of unintended persistence of potentially sensitive commercial data, file-path disclosure, and cross-session data exposure depending on how the host stores or shares local artifacts.

Ssd 3

Medium
Confidence
90% confidence
Finding
The skill instructs the agent to obtain API secrets directly from the user or read them from environment variables for later HTTP requests. Even though this is operationally common, embedding credential-collection guidance inside a skill increases the chance that secrets are solicited, handled, or exposed by the model layer rather than through a dedicated secret-management flow.

Static analysis

No suspicious patterns detected.