The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to perform geocoding and routing by executing shell commands where user-provided input (e.g., place names, addresses, and categories) is interpolated as arguments (e.g., python3 $MAPS search "<user_input>"). This pattern creates a potential for command injection if the execution environment does not adequately escape or sanitize the arguments.
[COMMAND_EXECUTION]: The workflow for 'Calibrate server time' involves executing system diagnostic utilities such as timedatectl status and chronyc sources to compare system time against geolocated timezone data.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes data from external, community-maintained APIs, which could theoretically contain malicious instructions designed to influence the agent's behavior.
Ingestion points: Data is fetched from OpenStreetMap (Nominatim), Overpass API, OSRM, and TimeAPI.io via the maps_client.py script (e.g., in the SKILL.md file).
Boundary markers: The documentation does not specify the use of delimiters or 'ignore instructions' warnings when processing these external data strings.
Capability inventory: The skill possesses terminal execution capabilities, file system access to its own scripts, and network access to public mapping APIs.
Sanitization: There is no evidence in the provided documentation that external API content is filtered or sanitized before being returned to the agent context.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to retrieve geographic and time data from well-known services including openstreetmap.org, project-osrm.org, and timeapi.io.

maps