Security audit

helword

Security checks for vulnerabilities and agentic risk

Overview

This is a low-impact hello-world greeting skill, with only a minor risk of activating on ordinary greetings.

This skill appears safe to install for a basic greeting/demo use case. Users should be aware it may activate on ordinary greetings or introduction requests because auto-invoke is enabled with broad trigger patterns.

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

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 (3)

Ae1

High

Category: analysis-evasion
Content: 2. Add a `SKILL.md` file with proper metadata
Confidence: 100% confidence
Finding: Referenced artifact was not completely inspected

Vague Triggers

Medium

Confidence: 96% confidence
Finding: The trigger patterns are broad enough to match routine conversation such as 'hi' or 'hey', which can cause this skill to auto-invoke unexpectedly. In an auto-invoked skill, that creates skill-routing collisions and may suppress or interfere with more relevant skills, even though the payload here is only a harmless greeting.

Vague Triggers

Low

Confidence: 87% confidence
Finding: The usage text says the skill activates when users 'introduce themselves', which broadens the apparent scope beyond simple greeting requests. That ambiguity can encourage over-triggering on normal conversational input and makes routing behavior less predictable, though the skill's actual action remains low impact.

Static analysis

No suspicious patterns detected.