ClawHub
Back to skill

Security audit

Baidu Web Search - 百度网页搜索

Security checks across malware telemetry and agentic risk

Overview

This skill is a straightforward Baidu web-search wrapper that sends user search queries to Baidu using a disclosed API key, with privacy considerations but no hidden or destructive behavior found.

Install this only if you are comfortable sending search terms to Baidu through your own BAIDU_API_KEY. Use a dedicated key if possible, monitor quota usage, and avoid searching secrets, confidential business information, or sensitive personal data. Treat the JisuAPI/JisuEPC mentions as optional suggestions, not required behavior.

SkillSpector

By NVIDIA
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
87% confidence
Finding
The trigger guidance is very broad and maps common phrases like '百度一下' or generic web-retrieval requests directly to this skill, which can cause over-invocation on ordinary user messages. In an agent setting, overly broad triggers increase the chance of unintended external requests, unnecessary disclosure of user queries to third-party services, and poor tool-selection safety boundaries.

Vague Triggers

Medium
Confidence
90% confidence
Finding
The invocation rules use ambiguous phrases such as '查一下 / 搜一下 / 最新消息' without constraints on subject matter or sensitivity, encouraging the agent to route many everyday requests into an external search workflow. This raises privacy and prompt-routing risk because casual user text, including sensitive or regulated topics, may be transmitted to Baidu unnecessarily.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.