Security audit

Web Search by Exa

Security checks across malware telemetry and agentic risk

Overview

This is a straightforward Exa web-search integration with privacy and API-key handling cautions, but no evidence of hidden or destructive behavior.

Install only if you are comfortable sending search queries, URLs, and optional profile/company research prompts to Exa. Protect any Exa API key, avoid putting secrets or unnecessary personal data in queries, and use people search only for authorized, lawful, appropriate research.

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 (2)

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: The skill explicitly promotes people search and LinkedIn/profile discovery but provides no warning or usage constraints around privacy-sensitive personal data, consent, or applicable legal/policy boundaries. In an agent context, this can normalize targeted profiling, candidate surveillance, or bulk personal-data collection without users recognizing the sensitivity of the operation.

Missing User Warnings

Low

Confidence: 79% confidence
Finding: The skill advertises URL crawling and clean content extraction but does not warn users that external page content will be fetched, processed, and returned to the model. This can lead to unintentional retrieval of third-party content, including sensitive or copyrighted material, and can expose users to prompt-injection or unsafe downstream handling if they assume the content is inert.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.