Security audit

IFQ App Builder

Security checks across malware telemetry and agentic risk

Overview

This skill is a prompt-bundle generator for app projects, and its disclosed file, shell, and downstream build behavior is coherent with that purpose.

Install only if you want an agent to generate app-build prompt bundles in your workspace. Before handing a generated bundle to another coding agent, review it because the next phase may install dependencies, run build tools, modify files, use network access, or require deployment credentials depending on the app you ask for.

SkillSpector

By NVIDIA

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 (7)

Vague Triggers

Medium

Confidence: 86% confidence
Finding: The trigger set for A-01 includes broad phrases like "desktop tool" that can match many ordinary user requests unrelated to this specific skill-routing context. In an agent system, overly broad routing can cause accidental invocation of a powerful scaffold-and-build skill, leading to prompt misselection, unintended code generation paths, or bypass of more appropriate specialized workflows.

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The trigger list contains ambiguous short terms such as "RN" and "Expo," which are likely to appear in many unrelated conversations and abbreviations. This increases the chance of accidental routing into the React Native/Expo template, which is risky because this skill produces executable build-ready prompt bundles intended for downstream coding agents.

Vague Triggers

Medium

Confidence: 84% confidence
Finding: The CLI route uses generic triggers like "CLI," "script," and "terminal tool," which overlap heavily with everyday software requests. In the context of a skill that turns brief user intent into end-to-end executable prompt bundles, generic activation terms can over-capture benign requests and steer them into an automated code-generation workflow the user did not intend.

Vague Triggers

Medium

Confidence: 91% confidence
Finding: The trigger list includes broad, common phrases such as "build app," "make app," and platform names that can match ordinary user requests beyond the narrowly intended three-sentence prompt-bundle workflow. This can cause unintended activation of the skill, leading the agent to route users into this skill in contexts where it is not the best or safest fit, increasing the chance of confused-deputy behavior or accidental execution of higher-permission skill capabilities.

Vague Triggers

Medium

Confidence: 81% confidence
Finding: The natural-language trigger example is broad enough that ordinary user requests could unintentionally activate this skill in hosts that perform loose semantic routing. Because this skill generates build-ready prompt bundles across many platforms, accidental invocation could cause the agent to prioritize this skill's instructions over the user's intended workflow, increasing prompt-scope and routing confusion.

Vague Triggers

Medium

Confidence: 79% confidence
Finding: Saying Claude Code loads the skill automatically when its description matches user intent leaves activation boundaries undefined, which can lead to over-broad matching and unintended skill execution. In an agent environment, ambiguous routing is security-relevant because a skill can inject workflow constraints, templates, or assumptions into sessions that did not explicitly request it.

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: The quickstart explicitly instructs users to hand the generated prompt bundle to an agent to scaffold, build, and run acceptance steps, which can lead to code execution, package installation, network access, and filesystem modification. In a skill whose purpose is to generate build-ready app prompts for multiple platforms, this omission is more dangerous because users are primed to treat the next step as routine and safe, despite potentially high-impact actions by the downstream agent.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.