Security audit

explore-vietnam

Security checks for vulnerabilities and agentic risk

Overview

This Vietnam flight-search skill is mostly purpose-aligned, but it can automatically install and run a global third-party CLI while sending itinerary details to an external service with limited disclosure.

Review before installing. Use this skill only if you trust the flyai npm package and are comfortable with a global CLI being installed and with your flight search details being sent to flyai or its booking provider. Prefer manually installing or approving the CLI yourself, and treat booking links as third-party output.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
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

Findings (5)

Missing User Warnings

Medium

Confidence: 96% confidence
Finding: The skill instructs the agent to run system commands and, if needed, globally install an npm package without requiring explicit user consent or warning that the host environment will be modified. In agent environments, this can lead to unapproved code execution, supply-chain exposure, and persistent changes to the machine running the skill.

Missing User Warnings

Medium

Confidence: 90% confidence
Finding: The skill is designed to send origin, destination, and travel dates through an external CLI/service but does not disclose this data flow or obtain user awareness/consent. Travel details are often sensitive personal information, and silent transmission to a third-party service creates privacy and compliance risk.

Vague Triggers

Medium

Confidence: 91% confidence
Finding: The trigger phrases "cheap" and "budget" are generic travel terms and can appear in many unrelated user utterances, making unintended playbook activation plausible. In a booking skill, this can cause the agent to run a flight-search flow when the user was discussing price preferences more generally, leading to incorrect tool use, confusion, or unwanted queries.

Vague Triggers

Medium

Confidence: 93% confidence
Finding: The phrases "fast" and "quick" are highly ambiguous and commonly used in ordinary conversation, so they do not safely identify a flight-search intent on their own. This can misroute the agent into the fastest-route playbook during unrelated exchanges, especially in a broad travel assistant that supports many services beyond flights.

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The trigger "direct" is too broad because it may refer to communication style, routing, or other non-flight concepts rather than nonstop airfare. In this multi-function travel skill, such ambiguity increases the chance of unintended direct-flight searches and incorrect downstream actions.

Static analysis

No suspicious patterns detected.