Security audit

trade-fair

Security checks across malware telemetry and agentic risk

Overview

This skill is mainly a real-time flight-search helper, but it needs review because it tells the agent to install and run an unpinned global CLI without clear user approval.

Install only if you trust the flyai CLI publisher and are comfortable with a global npm package being added to your machine. Require confirmation before any install or command execution, review the exact flight-search command, and treat origin, destination, travel dates, and preferences as information that may be sent to the travel provider.

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 (5)

Missing User Warnings

Medium

Confidence: 97% confidence
Finding: The skill directs the agent to run `npm i -g @fly-ai/flyai-cli` automatically if the CLI is missing, which performs a global package installation and executes external supply-chain code without explicit user consent. In an agent setting, this expands the trust boundary from a local tool invocation to arbitrary package retrieval and installation, creating risk of malicious packages, typosquatting, or unexpected system modification.

Missing User Warnings

Low

Confidence: 95% confidence
Finding: The skill mandates external CLI execution and even environment checks/install steps without clearly warning the user that their request will trigger local command execution and possible network access. In an agent environment, lack of disclosure is dangerous because users may not realize the agent is crossing from text generation into tool execution with side effects and external data transfer.

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The trigger set for the cheapest-option playbook includes very broad terms like "cheap" and "budget", which can match many unrelated travel conversations and cause the wrong booking workflow to activate. In a transactional travel skill, unintended invocation can steer users toward an unintended search mode and produce misleading or undesired booking results.

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The fastest-route playbook is triggered by generic words like "fast" and "quick", which are highly ambiguous and may appear in many requests unrelated to flight duration. This can cause incorrect playbook selection and inaccurate automation behavior in a booking flow, especially when users are discussing response speed or general urgency rather than itinerary preference.

Vague Triggers

Medium

Confidence: 84% confidence
Finding: The fallback condition "0 results from above playbooks" is underspecified and may activate broad search behavior without clear guardrails on when prior playbooks were attempted or how failure is determined. Because the fallback includes a free-form keyword search, accidental or overly broad activation can widen query scope and return irrelevant or unintended results, reducing reliability and potentially mishandling user intent.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.