ClawHub
Back to skill

Security audit

Search Agent Skill

Security checks across malware telemetry and agentic risk

Overview

This skill does not appear malicious, but it needs review because it is advertised as a real AI search and fact-checking tool while the implementation returns placeholder search results.

Review before installing or relying on this skill. The main risk is not hidden malware behavior; it is that the package appears demonstration-quality while claiming real search, AI summarization, fact checking, source filtering, and credibility scoring. Do not use it for sensitive queries or decisions unless the publisher replaces the placeholder search logic with real, transparent providers and documents what data is sent where.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The skill is explicitly designed to send user queries to external search APIs, fetch third-party webpages, and aggregate results, yet the user-facing description does not clearly warn that prompts may be transmitted to outside providers and websites. This can expose sensitive user queries, internal project names, credentials pasted by mistake, or other confidential context to third parties without informed consent.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.env_credential_access

Environment variable access combined with network send.

Critical
Code
suspicious.env_credential_access
Location
index.js:14