ClawHub
Back to skill

Security audit

抖音-蓝海爆品采集师 | 青虎AI

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed Douyin and 1688 product-research skill that uses Qinghu API data with user authorization, with only minor trigger-breadth cautions.

Install only if you intend to use Qinghu's Douyin/1688 research APIs. Expect to provide or expose a Qinghu API token, approve API calls that may consume Qinghu credits, and receive exported table files for larger datasets.

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
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill description defines activation across very common product-research intents such as finding trending products, niche demand, keyword popularity, and sourcing. Because these intents overlap with many ordinary commerce requests, the skill can trigger when the user did not specifically ask for this capability, causing tool overreach and unnecessary access to external paid APIs and export flows. The billing/authorization safeguards reduce direct harm, but broad invocation still increases the chance of inappropriate routing and unintended data operations.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The listed trigger phrases are short, generic, and lack boundaries, for example requests about finding niche products, checking whether a term is popular, or helping find a supplier. In a multi-skill environment, these phrases are likely to collide with unrelated research, sourcing, or social-media analysis tasks, which can misroute conversations into this skill and lead to unnecessary credential prompts, paid tool calls, or file exports. The skill context makes this more concerning because it is user-invocable and connected to authenticated third-party data sources.

Content

No source excerpt is available for this finding.