The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the ax CLI to manage administrative resources on the Arize platform. It incorporates a safety-first approach by requiring the agent to use AskUserQuestion for confirmation before executing destructive commands like delete or revoke.
[EXTERNAL_DOWNLOADS]: The documentation includes instructions for installing the arize-ax-cli package using standard package managers such as pip, pipx, and uv. This is the official CLI tool provided by the vendor for interacting with their service.
[CREDENTIALS_UNSAFE]: The skill includes explicit security constraints that prevent the agent from reading .env files or searching the filesystem for credentials. It mandates that API keys should only be referenced via environment variables and never echoed, logged, or requested in plain text from the user.
[PROMPT_INJECTION]: The skill is designed to process output from the Arize API (via the CLI). While this presents a surface for indirect prompt injection if an attacker controls resource names, the skill mitigates this risk by requiring human confirmation for all high-impact actions.

arize-admin