The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local Node.js script (node dist/index.js) to perform searches and manage authentication. This is standard behavior for a CLI-based skill and uses localized paths.
[DATA_EXFILTRATION]: Communication is restricted to the official vendor domain https://zizo.pro. There is no evidence of sensitive local data being transmitted to external servers.
[CREDENTIALS_UNSAFE]: The skill implements strong credential safety patterns. It explicitly warns against hardcoding tokens and provides a secure local configuration mechanism (~/.config/zizo/config.json) with restricted file permissions (0600) via an interactive login flow.
[INDIRECT_PROMPT_INJECTION]: The skill correctly identifies and labels search results (titles, descriptions, URLs) as untrusted third-party content. It includes specific security notices advising agents to treat this data as untrusted and avoid passing it into privileged operations without sanitization.
[EXTERNAL_DOWNLOADS]: The skill does not perform any unverified external downloads or remote script execution. All functionality is contained within the distributed package and official API endpoints.

zizo