Security audit

Strategic Mentality

Security checks across malware telemetry and agentic risk

Overview

The skill is mostly a disclosed business strategy coach, but it needs review because its metadata lists unrelated financial-style capabilities and its recurring Telegram/accountability flows lack clear opt-in, retention, and disable controls.

Review before installing. Only enable the Telegram heartbeat after confirming the destination chat, what business metrics will be sent, where scores are stored, how long they are kept, and how to pause or delete them. Do not grant crypto, purchasing, payment, CRM, email-sending, scraping, or calendar authority unless a separate workflow clearly needs it and requires user approval for each external action.

SkillSpector

By NVIDIA

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (5)

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: The skill defines recurring Telegram outreach and explicitly stores weekly performance scores for trend tracking, but provides no notice, consent flow, retention limit, or handling guidance for that user data. Because the data includes behavioral performance history and business outcomes, this creates a real privacy and data-governance issue rather than a harmless omission.

Missing User Warnings

Medium

Confidence: 88% confidence
Finding: The skill prompts the user to provide sensitive business metrics such as revenue, leads, calls, and deals closed, then states it will calculate scores and perform trend analysis without any privacy warning or data-handling constraints. In business-strategy context, these metrics can be commercially sensitive, so collecting and analyzing them without transparency increases exposure if logs, chats, or downstream systems are accessed or reused.

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The trigger language is very broad and includes generic phrases like 'apply a framework' or 'evaluate this opportunity,' which can cause the skill to activate in situations far beyond its intended scope. Unintended invocation can override more appropriate skills, inject business/sales framing into unrelated requests, and increase the chance that auxiliary behaviors like scheduled accountability or context overlays are surfaced when not relevant.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The file explicitly recommends automated scoring from scraper/API data and proposes checks against websites, social activity, and business categories, but provides no privacy notice, lawful-basis guidance, data-minimization limits, or warning against collecting personal data. In a business prospecting context, this can normalize bulk profiling and data collection workflows that may process personal or sensitive business-contact information without appropriate safeguards.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The guidance repeatedly uses directives like 'ATTACK', recommends targeting competitor weaknesses, and explicitly suggests 'scraping tools = reconnaissance' without any caution about legality, terms-of-service compliance, privacy, or ethical limits. In a business strategy skill, this can normalize improper competitor targeting and unauthorized data collection, increasing the chance a user operationalizes risky conduct.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.