The Agent Skills Directory

[SAFE]: The skill defines rules for preventing redundant tool calls and ensuring accurate parameter passing. No malicious behavior or security bypass attempts were detected.
[INDIRECT_PROMPT_INJECTION]: The skill handles data from tool observations.
Ingestion points: Processes latitude and longitude from Geocoding tool outputs (referenced in rules.md).
Boundary markers: Instructions explicitly tell the agent to copy coordinates from <obs> tags.
Capability inventory: No dangerous functions like subprocess, network access, or file writes are present in the skill.
Sanitization: The validator script enforces strict parameter matching against a defined schema.
[DYNAMIC_EXECUTION]: The validator.py script utilizes an internal platform library to perform static checks on tool parameters. It does not execute arbitrary code or use untrusted inputs in an unsafe manner.

Geocoding observation to coordinate service