The Agent Skills Directory

[COMMAND_EXECUTION]: The skill orchestrates multiple shell commands using the Treasure Data CLI (tdx), curl, and jq. These operations are essential for its primary function: discovering data schemas, configuring real-time infrastructure, and deploying personalization entities.
[EXTERNAL_DOWNLOADS]: The skill performs network operations targeting official Treasure Data endpoints, such as api-cdp.treasuredata.com and p13n.in.treasuredata.com. These are well-known vendor services required for the skill to interact with the platform.
[CREDENTIALS_UNSAFE]: The workflow manages sensitive credentials by instructing users to set the TD_API_KEY as an environment variable and subsequently using it in authorized headers (Authorization: TD1 ${TD_API_KEY}). This is a standard and safe administrative practice for CLI-based automation tools.
[PROMPT_INJECTION]: Instructions within the skill are focused on operational guidance and validation logic. There are no attempts to override safety filters or bypass agent constraints.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from the Treasure Data environment (such as table schemas and configuration objects). While this constitutes an attack surface, the skill implements validation steps (referencing an external validation skill) to ensure data integrity and prevent API errors, which is a recommended mitigation strategy.

rt-setup-personalization