AI Opportunity Scout

• Outcome: Pass
• Latest audit: May 28, 2026 at 11:35 AM UTC
• Version:
  1.0.0

Overview

This is a coherent market-research helper that searches public sources and writes a local opportunity report, with command-safety and privacy cautions but no evidence of hidden or destructive behavior.

Use this for public market research. Do not include confidential plans, customer names, secrets, or private business details in the niche prompt, and only let it run local CLI tools you trust; agents should quote or pass niche text as structured arguments rather than building raw shell strings.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration

Findings (3)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 87% confidence
• Finding: The skill instructs use of local scripts and generates files (`scripts/scout.py`, `opportunities.json`, `report.md`) but does not declare the corresponding file read/write capabilities. Undeclared capabilities create a transparency and policy-enforcement gap: a reviewer or runtime may assume the skill is read-only/research-only when it can access or modify local workspace data.

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 96% confidence
• Finding: The declared behavior says the skill scans external sources and performs market discovery, but the implementation described in the file primarily relies on local script execution and templated report generation without guaranteed evidence collection. This mismatch is dangerous because it can mislead users and orchestrators about what data is actually gathered and how results are produced, enabling fabricated or unsupported outputs and masking additional local execution behavior.

Vague Triggers

Medium

• Confidence: 78% confidence
• Finding: The trigger phrases are broad enough that the skill may activate for generic market research or product questions outside the user's intent. Over-broad invocation can cause unnecessary tool use, file operations, or external searches, increasing the chance of unintended data access or noisy autonomous behavior.

VirusTotal

66/66 vendors flagged this skill as clean.