Security audit

hormozi-business-kickstart

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed business-manual generator with local PDF/DOCX conversion scripts, though users should be careful about sharing confidential business or demographic details.

Install only if you want an agent to ask detailed business-strategy questions and create local Markdown/PDF/DOCX outputs. Avoid providing unnecessary personal, protected demographic, regulated, or highly confidential financial data, and ensure generated files are written to intended locations.

SkillSpector

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
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

Findings (7)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 79% confidence
Finding: The skill documents companion files and activation behavior that require reading local files such as `SKILL_PROMPT.md` and `QUESTIONNAIRE.md`, yet no corresponding permissions are declared. Undeclared file-read capability weakens transparency and policy enforcement because the runtime may access files the user and platform did not explicitly authorize.

Tp4

High

Category: MCP Tool Poisoning
Confidence: 86% confidence
Finding: The description presents a narrowly scoped business-manual generator, but the documented scripts are generic Markdown-to-PDF/DOCX converters with broad formatting controls. That mismatch can hide a more general document-processing capability than users expect, increasing the risk of misuse, unsafe content handling, or policy bypass through arbitrary Markdown conversion.

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: The questionnaire explicitly solicits potentially sensitive business and demographic information such as income bracket, company revenue, geography, age, gender, psychographics, budgets, CAC/LTV, and lead-source data without any accompanying notice about sensitivity, minimization, retention, or safe handling. In a business-strategy skill this collection may be relevant, but absent user-facing warning and privacy guidance it creates unnecessary exposure of confidential commercial data and personal data that could be mishandled, over-collected, or shared downstream.

Vague Triggers

High

Confidence: 89% confidence
Finding: Broad triggers like 'help me get more customers' or 'I need more leads' can cause the skill to activate for ordinary business conversations well outside the intended scope. Overbroad activation increases the chance of accidental prompt takeover, inappropriate file loading, and unintended execution of the skill's full business-strategy workflow when a simpler response was expected.

Vague Triggers

Medium

Confidence: 83% confidence
Finding: The activation guidance says to trigger on listed phrases but does not define exclusions, confidence thresholds, or fallback behavior for partial matches. This ambiguity makes unintentional activation more likely, which is dangerous in a skill that loads additional prompt files and drives a structured workflow that may override the user's immediate intent.

Vague Triggers

Medium

Confidence: 83% confidence
Finding: The trigger condition is broad enough that normal user conversation containing phrases like 'grand slam offer' or related terms could invoke the skill unintentionally. In an agent environment, unintended activation can cause unauthorized file generation, execution of downstream workflow steps, or accidental use of local tooling without clear user intent.

Missing User Warnings

Low

Confidence: 87% confidence
Finding: The prompt instructs the agent to run local conversion scripts and deliver generated files, but does not require user-facing disclosure or confirmation before executing those local actions. In agentic systems, undisclosed tool execution increases the risk of unsafe file handling, misuse of local resources, and surprising side effects, especially if filenames, paths, or delivery channels are influenced by user input elsewhere.

Static analysis

No suspicious patterns detected.