The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The scripts use npx to fetch and run the @toolbox-sdk/server@1.1.0 package. This is a standard and legitimate method for executing versioned utility tools in this environment.- [COMMAND_EXECUTION]: Core functionality is executed by spawning the npx command using child_process.spawn. The implementation follows platform conventions for invoking CLI tools with provided parameters.- [SAFE]: Environment variables and secrets are managed by loading from local .env files, which is an established and safe practice for secret management in this context.- [PROMPT_INJECTION]: The skill processes user-provided SQL and search prompts as part of its primary data management functionality. While these inputs represent a theoretical surface for indirect prompt injection, they are necessary for the skill's intended purpose.

bigquery-data