The Agent Skills Directory

[COMMAND_EXECUTION] (SAFE): The skill executes a local script via bun run to provide its interactive interface. This is restricted to its own source directory and contains no dynamic or unsanitized inputs from external sources in the command string itself.
[INDIRECT_PROMPT_INJECTION] (LOW): The skill processes flight information which may originate from external APIs. While this is an ingestion point for untrusted data, the skill's role is primarily visual selection, and it lacks high-privilege capabilities that would allow such data to execute harmful actions.
[DATA_EXPOSURE] (SAFE): No hardcoded credentials, API keys, or attempts to access sensitive system files were identified.

flight