ClawHub
Back to skill

Security audit

Geoskill: Impervious Surface Mapping

Security checks across malware telemetry and agentic risk

Overview

This skill mostly matches its mapping purpose, but its remote-download path can substitute synthetic demo data for real imagery, so users should review it before relying on results.

Install only if you are comfortable with a CLI that reads local geospatial files, writes outputs under the chosen output directory, and may contact Microsoft Planetary Computer when bbox/date inputs are used. Treat remote-download results carefully: check the manifest and console output to confirm the analysis used real imagery rather than synthetic demo data, and consider pinning dependencies for controlled environments.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (9)

Lp3

Medium
Category
MCP Least Privilege
Confidence
93% confidence
Finding
The skill documentation describes file-reading and file-writing behavior via raster inputs, GeoJSON inputs, and output artifacts, but it does not declare permissions for those capabilities. Undeclared filesystem access is dangerous because users and orchestration systems may grant or assume a lower privilege profile than the skill actually requires, reducing transparency and weakening security controls around local data access.

Description-Behavior Mismatch

Medium
Confidence
90% confidence
Finding
The script expands beyond local impervious-surface estimation by modifying import paths and optionally enabling external data discovery/download through a shared fetcher. That broader capability can trigger unanticipated network access and pull in external code/data paths, which is a real security and trust-boundary issue even if intended as a convenience feature.

Description-Behavior Mismatch

Medium
Confidence
95% confidence
Finding
When auto-download fails or returns a single-band asset, the tool silently falls back to synthetic data and still produces outputs, which can mislead downstream users into believing the results correspond to the requested AOI/date. In geospatial decision workflows, this integrity failure can cause incorrect analysis, reporting, or operational decisions.

Missing User Warnings

Low
Confidence
88% confidence
Finding
The skill can automatically download remote Sentinel-2 data from Microsoft Planetary Computer, but the main description does not clearly warn users that network access and external data transfer may occur. This matters because remote fetching can expose user-supplied locations and dates to a third party, introduce supply-chain and availability risks, and surprise users operating in restricted or offline environments.

Missing User Warnings

Low
Confidence
84% confidence
Finding
The auto-download path sends user-supplied bbox/AOI and date-range parameters to an external service without an explicit user-facing privacy warning or confirmation. While expected for remote data retrieval, this can disclose sensitive locations or project timing information in contexts where AOIs are confidential.

Unpinned Dependencies

Low
Category
Supply Chain
Content
geoskill-data-fetcher>=0.2.0

# Third-party dependencies used by this skill
numpy
rasterio
shapely
Confidence
95% confidence
Finding
numpy

Unpinned Dependencies

Low
Category
Supply Chain
Content
# Third-party dependencies used by this skill
numpy
rasterio
shapely
Confidence
95% confidence
Finding
rasterio

Unpinned Dependencies

Low
Category
Supply Chain
Content
# Third-party dependencies used by this skill
numpy
rasterio
shapely
Confidence
94% confidence
Finding
shapely

Unpinned Dependencies

Low
Category
Supply Chain
Content
# Requirements for impervious-surface-mapping
# Shared data-download library (Planetary Computer / NASA POWER / OSM)
geoskill-data-fetcher>=0.2.0

# Third-party dependencies used by this skill
numpy
Confidence
96% confidence
Finding
geoskill-data-fetcher>=0.2.0

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.