ClawHub
Back to skill

Security audit

Linkedin Engager Analytics

Security checks across malware telemetry and agentic risk

Overview

This skill is mostly transparent, but it can scrape and profile LinkedIn engagers for outreach and its allowed scope is inconsistent.

Install only if you intend to analyze LinkedIn posts you own or have clear permission to track. Before running it, provide explicit post URLs, cap the number of engagers, and avoid competitor-audience scraping or same-day outreach unless you have a lawful and platform-compliant basis.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (4)

Description-Behavior Mismatch

Medium
Confidence
94% confidence
Finding
The skill scope expands from analyzing engagement on the user's own posts to reviewing competitor or niche posts, which materially increases the chance of non-consensual audience scraping and profiling of third parties. In this context, the skill is specifically designed to segment and operationalize outreach against those engagers, so the scope creep makes privacy, platform-abuse, and harassment risks more acute.

Intent-Code Divergence

Medium
Confidence
83% confidence
Finding
The documentation narrows one form of tracking in the description but later introduces cross-account activity checks via recent-comment history, creating a misleading boundary around surveillance behavior. That inconsistency can cause users or downstream agents to believe the skill is safer or narrower than it really is, increasing the risk of unintended tracking and broader collection of activity data.

Vague Triggers

Medium
Confidence
86% confidence
Finding
The trigger phrase "audience analytics" is generic and could collide with unrelated requests about other platforms, campaigns, or internal reporting. In a skill that can scrape LinkedIn engager data and produce outreach actions, such ambiguity raises the chance of over-collection and unintended processing of third-party personal or professional information.

Vague Triggers

Medium
Confidence
86% confidence
Finding
The trigger phrase "audience analytics" is generic and could collide with unrelated requests about other platforms, campaigns, or internal reporting. In a skill that can scrape LinkedIn engager data and produce outreach actions, such ambiguity raises the chance of over-collection and unintended processing of third-party personal or professional information.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.