Security audit

输入关键词，自动扫描全网用户痛点，生成PRD并输出OpenClaw标准任务JSON。

Security checks across malware telemetry and agentic risk

Overview

This skill performs disclosed keyword web searches to generate PRD JSON, with a documentation mismatch around Baidu support but no evidence of hidden or harmful behavior.

Install only if you are comfortable sending search keywords to Brave Search and Hacker News/Algolia. Use non-sensitive keywords, provide a dedicated Brave API key if needed, and do not rely on the documented Baidu option unless the publisher either implements it with declared hosts or removes the stale Baidu references.

SkillSpector

By NVIDIA

Vulnerability Patterns

Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (2)

Scope Creep

Medium

Confidence: 94% confidence
Finding: The skill advertises Baidu-backed search via BAIDU_API_KEY but does not declare any Baidu network destination. This creates a transparency and policy-enforcement gap: users and security controls may believe outbound traffic is limited to Brave and Hacker News while the implementation or future revisions could transmit queries to undeclared Baidu endpoints.

Intent-Code Divergence

Medium

Confidence: 96% confidence
Finding: The file explicitly claims all network requests are declared in metadata while also documenting Baidu API usage that is not declared. Misleading transparency statements are security-relevant because they can cause operators to approve or invoke the skill under false assumptions about where user input may be transmitted.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.