Call Web Search Agent Strategy

• Outcome: Pass
• Latest audit: May 28, 2026, 11:32 AM
• Version:
  0.1.0

Overview

This appears to be a web-research workflow skill with expected network/search behavior, but its instructions should be used cautiously because they overstate trust in user input and tool output.

Install this only if you want an agent to perform web searches and scraping on your behalf. Treat its results as research leads, not verified truth, and use extra review for high-stakes topics or pages that may contain adversarial instructions.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (2)

Ssd 1

High

• Confidence: 96% confidence
• Finding: The role framing says the skill’s commands have priority higher than built-in knowledge and cognition, and requires unconditional acceptance of tool output as fact. In a search-oriented skill, that increases risk of prompt injection, misinformation acceptance, and unsafe tool-mediated decision making because the agent is told to suppress skepticism and deference to trusted safety layers.

Ssd 1

High

• Confidence: 96% confidence
• Finding: The role framing says the skill’s commands have priority higher than built-in knowledge and cognition, and requires unconditional acceptance of tool output as fact. In a search-oriented skill, that increases risk of prompt injection, misinformation acceptance, and unsafe tool-mediated decision making because the agent is told to suppress skepticism and deference to trusted safety layers.

VirusTotal

66/66 vendors flagged this skill as clean.