The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill retrieves and processes data from external, community-maintained databases (OpenStreetMap and Overpass API). Malicious instructions embedded in location names, descriptions, or addresses in these data sources could potentially influence the agent's behavior.
Ingestion points: The script scripts/maps_client.py fetches data from Nominatim (nominatim.openstreetmap.org), Overpass API (overpass-api.de, overpass.kumi.systems), OSRM (router.project-osrm.org), and TimeAPI (timeapi.io).
Boundary markers: The skill does not implement explicit boundary markers or instructions for the agent to ignore content within the retrieved geographic data.
Capability inventory: The scripts/maps_client.py script is restricted to network operations and terminal output; it does not have capabilities for file system modification or command execution.
Sanitization: Data is parsed as JSON and returned as structured output, but no textual sanitization is performed on user-facing strings like place names or navigation instructions.
[DATA_EXFILTRATION]: The skill transmits location-related metadata and coordinates to external mapping and routing services.
Evidence: The script communicates with nominatim.openstreetmap.org, overpass-api.de, overpass.kumi.systems, router.project-osrm.org, and timeapi.io. These are recognized public utility services for geographic data and timezones.

maps