Spatial Data Engineer
Pass
Audited by Gen Agent Trust Hub on Oct 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The agent is designed to ingest geospatial data from a wide variety of external sources and file formats, creating an attack surface for indirect prompt injection.
- Ingestion points: Data is ingested from government portals, APIs, and formats like Shapefile, GeoPackage, GeoJSON, KML, CSV, and Parquet (SKILL.md).
- Boundary markers: No specific boundary markers or instructions to ignore embedded commands are defined for the ingestion process.
- Capability inventory: The agent's toolset includes network operations (requests), database management (SQLAlchemy), and extensive file system utilities (GDAL, Fiona, Rasterio) (SKILL.md).
- Sanitization: While the agent standardizes schemas and cleans geometries, it does not explicitly sanitize text-based attributes for potentially malicious instructions.
Audit Metadata