The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local script language-router.py with the --action supported-languages argument to retrieve supported language information and routing logic. This is a standard functional requirement for the skill.
[PROMPT_INJECTION]: The skill creates an ingestion surface for indirect prompt injection by allowing users to add arbitrary strings to a 'do-not-translate' (DNT) list stored in profile.json.
Ingestion points: User input via the Term parameter and loading of profile.json and agency SOPs.
Boundary markers: No explicit markers are defined for the storage or interpolation of DNT terms.
Capability inventory: The skill writes to local files (profile.json) and executes a local script (language-router.py).
Sanitization: The skill performs JSON integrity checks and basic string validation but does not sanitize terms for instruction-like content that could influence downstream agents.

language-config