Security audit

peekaboo

Security checks for vulnerabilities and agentic risk

Overview

This skill clearly describes a macOS UI automation tool with powerful but purpose-aligned screen, input, clipboard, and app-control capabilities.

Install only if you are comfortable granting a tool macOS Screen Recording and Accessibility access. Avoid using it around passwords, tokens, private documents, or sensitive clipboard contents unless the task requires it, and do not hardcode real credentials in automation commands or scripts.

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
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (3)

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: The skill explicitly advertises clipboard read/write support, which can expose highly sensitive data such as passwords, tokens, personal data, and copied files. In a UI automation skill, documenting this capability without any privacy warning or usage guardrails increases the chance that agents or users will access clipboard contents inappropriately or overlook the sensitivity of that channel.

Ssd 3

Medium

Confidence: 95% confidence
Finding: The documented workflow walks through entering login credentials into a form, which directly encourages handling secrets via UI automation. In this skill's context, that is especially risky because the tool also supports screen capture, annotation, and scripting, creating multiple opportunities for credentials to be recorded, cached, replayed, or exposed beyond the intended target application.

Missing User Warnings

Medium

Confidence: 96% confidence
Finding: The example uses a password-like string in an automation flow, normalizing the practice of entering secrets through UI automation without any warning about secret handling. This is dangerous because automation layers may log commands, preserve snapshots, expose typed values in transcripts, or encourage unsafe patterns for real credentials.

Static analysis

No suspicious patterns detected.