Security audit

Militia Training (EN)

Security checks for vulnerabilities and agentic risk

Overview

This skill is markdown-only and does not access your system, but it can steer an AI into giving operational militia, firearms, and combat-planning guidance.

Install only if you intentionally want an agent to discuss militia-style military training. Keep use limited to lawful, authorized, high-level educational or safety contexts, and avoid using it for real-world weapons employment, target engagement, ambush or raid planning, patrol execution, or medical treatment decisions.

SkillSpector

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
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 (3)

Natural-Language Policy Violations

High

Confidence: 98% confidence
Finding: This file provides ready-to-use prompt templates for anti-aircraft recognition, NBC defense, shooting instruction, combat patrol planning, and battlefield first aid. Because the templates are framed as a general-purpose assistant skill and include operationally useful military guidance such as patrol orders, actions on enemy contact, fire coordination, and weapons training, they materially enable combat capability rather than limiting themselves to high-level historical or safety-focused education.

Natural-Language Policy Violations

High

Confidence: 98% confidence
Finding: This section contains actionable weapons-use and small-unit combat guidance, including shot sequence, zeroing, patrol formations, actions on contact, and suppressive fire. That conflicts with the file's own safety policy to decline real-world violence instruction and materially increases the chance the skill can facilitate operational violent activity rather than merely provide historical or abstract education.

Natural-Language Policy Violations

High

Confidence: 95% confidence
Finding: The chemical response section provides specific agent-type antidote/treatment instructions, which goes beyond general defensive guidance and contradicts the stated rule to decline NBC agent specifics. Even if framed as defense, medication recommendations for nerve, blister, and blood agents can be unsafe, medically inappropriate, or repurposed in harmful contexts.

Static analysis

No suspicious patterns detected.