The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a bash script to handle request logic and a Python script to validate input types and structure. These operations are restricted to the local environment for the purpose of the skill's primary function.
[EXTERNAL_DOWNLOADS]: The skill performs an outbound network request using curl to ai-factory.frevana.com to fetch search results. This is the core functionality described in the documentation and targets the vendor's own infrastructure.
[CREDENTIALS_SAFE]: The skill handles authentication via the FREVANA_TOKEN environment variable or an explicit --token flag. It does not hardcode credentials and appropriately prompts for input if the token is missing in an interactive session.
[DYNAMIC_EXECUTION]: Python is used as a sub-process to securely parse and validate JSON payloads, avoiding unsafe evaluation of user input.

google-ai-mode