Security audit

Structs Reconnaissance

Security checks for vulnerabilities and agentic risk

Overview

This skill is a game reconnaissance helper that runs read-only Structs lookups and saves the resulting notes in clearly named local intel files.

Install only if you are comfortable with the agent keeping local Structs intelligence notes under memory/intel. Review or delete those files when you no longer want retained scouting records, especially if they include sensitive strategy or opponent assessments.

A.I.G

Vulnerability Patterns

Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

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
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (3)

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The skill explicitly instructs persistent writes to memory/intel without warning the user that it will modify local files. In an agent setting, silent persistence can create privacy, integrity, and workspace-tampering risks, especially when the collected data concerns competitive targeting and may be retained longer than intended.

Intent-Code Divergence

Medium

Confidence: 83% confidence
Finding: The manifest and surrounding procedure describe this skill as gathering intelligence via read-only queries and writing findings to memory/intel/. Line L053 introduces an 'after action' verification step to confirm state changes, implying use alongside state-changing operations that are not otherwise part of this skill's documented behavior.

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The instructions to create directories and update dossier files authorize filesystem changes without any safety notice, approval checkpoint, or path constraints beyond the intended folder names. In an automated environment this can lead to unreviewed local state changes, accidental overwrites, or accumulation of sensitive intelligence artifacts on disk.

Static analysis

No suspicious patterns detected.