ClawHub
Back to skill

Security audit

flights-overhead

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed flight-lookup skill, but users should confirm the location source before coordinates are sent to flight services.

Before installing or using this skill, expect it to use your location and send an approximate coordinate area to OpenSky. Prefer device location or an explicitly provided city/area, and ask the agent not to read USER.md or perform route web searches unless you want that extra detail.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Context-Inappropriate Capability

Medium
Confidence
91% confidence
Finding
The skill instructs the agent to read a home address from USER.md as a fallback, which expands data access beyond what is minimally necessary for an overhead-flight query. Even if operationally useful, this creates unnecessary exposure of sensitive personal location data and can normalize cross-context access to profile information for a simple lookup task.

Context-Inappropriate Capability

Medium
Confidence
85% confidence
Finding
The route-lookup step instructs the agent to perform web searches unrelated to the core function of identifying nearby aircraft from OpenSky data. This broadens tool use and data flow, increases prompt-injection exposure from search results, and can reveal more about observed flights than the user requested.

Context-Inappropriate Capability

Medium
Confidence
82% confidence
Finding
Referencing a separate flight_tracker.py tool for enrichment extends the skill beyond its declared purpose and introduces hidden dependency/tooling risk. Additional tools may access different datasets or permissions, making the effective behavior of the skill broader than users and reviewers expect.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.