arcgis-layers

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of data from various external file formats and remote URLs, creating a surface for potential indirect prompt injection.
  • Ingestion points: GeoJSONLayer, CSVLayer, ParquetLayer, GeoRSSLayer, and StreamLayer (in SKILL.md) are documented to load content from external URLs.
  • Boundary markers: The instructions do not specify the use of boundary markers or instructions for the agent to ignore content within these data streams.
  • Capability inventory: The skill focus is limited to mapping and visualization; it does not include capabilities for shell execution, filesystem modification, or system persistence.
  • Sanitization: No explicit data sanitization or schema validation processes are mentioned in the documentation for handling these external data sources.
  • [EXTERNAL_DOWNLOADS]: Fetches geographic data and map configurations from well-known services.
  • The documentation includes examples fetching GeoJSON data from the USGS (earthquake.usgs.gov) and map services from official ArcGIS Online endpoints (services.arcgis.com, basemaps.arcgis.com).
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 08:38 AM