Security audit

Airpoint 1.3.16

Security checks for vulnerabilities and agentic risk

Overview

This skill is openly a Mac desktop-control assistant, but it gives broad screen-reading and click/type authority without enough user-facing boundaries for sensitive actions.

Install only if you are comfortable granting Airpoint broad ability to view your screen and control your Mac. Avoid exposing passwords, private messages, financial data, or admin prompts while using it, and supervise tasks closely, especially when using `--no-wait` or asking it to interact with email, chat, browser sessions, or system settings.

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

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

Findings (2)

Missing User Warnings

High

Confidence: 98% confidence
Finding: The documentation describes an agent with screen-reading, keyboard/mouse control, screenshot capture, and autonomous workflow execution, but omits an explicit warning about the privacy and system-impact risks of granting those capabilities. Because the skill can observe sensitive on-screen content and manipulate any accessible application, users may underestimate the danger and authorize actions affecting credentials, personal data, communications, or system configuration.

Vague Triggers

Medium

Confidence: 95% confidence
Finding: The skill advertises unconstrained natural-language control of a Mac, including opening apps, reading the screen, typing, and performing multi-step autonomous actions, without defining narrow trigger boundaries or safety gating. In practice, this enables highly privileged, user-facing automation that could be invoked for sensitive actions across arbitrary applications, increasing the risk of privacy violations, destructive system changes, or unintended exfiltration if the skill is used loosely or prompted unsafely.

Static analysis

No suspicious patterns detected.