The Agent Skills Directory

[DATA_EXFILTRATION] (SAFE): The skill is designed to manage Google API credentials locally. Files such as credentials.json and token.json are used for storing OAuth secrets; however, the provided versions use placeholders like YOUR_CLIENT_ID and YOUR_ACCESS_TOKEN, ensuring no sensitive data is pre-exposed.
[PROMPT_INJECTION] (LOW): The skill presents a surface for indirect prompt injection. Because it retrieves and displays data from external sources (Google Analytics and Tag Manager), a malicious actor with control over the metadata in those accounts (e.g., account names or tag descriptions) could attempt to embed instructions to influence the agent. This is a common risk for tools that read untrusted external data.
[COMMAND_EXECUTION] (SAFE): The skill uses the Bash tool specifically to execute its internal Node.js scripts. The scripts use a custom argument parser and do not exhibit obvious command injection vulnerabilities.

gapi