The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The geocoder.py script communicates with external geocoding service providers (Nominatim, Google, Bing, and ArcGIS) to perform address-to-coordinate lookups and reverse geocoding. These are established, well-known services required for the skill's core functionality.
[INDIRECT_PROMPT_INJECTION]: The skill provides an attack surface for indirect prompt injection as it ingests untrusted external data in the form of unstructured address strings and geographic files (CSV, KML, GeoJSON) provided by the user.
Ingestion points: address_parser.py, geocoder.py, geo_visualizer.py, and territory_mapper.py all read and process user-supplied files or text input.
Boundary markers: The instructions do not define specific delimiters for separating user data from agent instructions.
Capability inventory: The skill can perform network requests to geocoding APIs and write output to the local file system (HTML maps, CSV results).
Sanitization: The scripts use standard parsing libraries like pandas and geopandas which provide basic structure validation, though they do not explicitly sanitize for natural language instructions embedded within data fields.
[COMMAND_EXECUTION]: The skill consists of multiple Python scripts designed to be executed via the command line with various arguments for processing files and data. This is standard behavior for a utility-based toolkit.

geo-toolkit