The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized data exfiltration were detected.
[DATA_EXPOSURE]: The skill manages a sensitive Yandex OAuth token. It stores this token in a local .env file within the skill directory. This is standard practice for local CLI tools, and the documentation appropriately advises users on how to obtain and store it.
[COMMAND_EXECUTION]: The shell scripts use standard utilities like curl for API communication and jq for JSON processing. All command-line arguments are handled in a controlled manner, specifically utilizing jq to build JSON payloads and sanitize input into appropriate formats (like numbers).
[EXTERNAL_DOWNLOADS]: All network requests are directed to official Yandex domains (api.direct.yandex.com, api-sandbox.direct.yandex.com, and oauth.yandex.ru) required for the skill's primary purpose of API management.

yandex-direct