The Agent Skills Directory

[SAFE]: The skill demonstrates high security awareness by including explicit guardrails for the agent. It strictly forbids searching the filesystem for secrets and provides clear instructions on using environment variables for credential management.- [COMMAND_EXECUTION]: The skill uses the official ax CLI tool to perform its operations, such as creating evaluators and managing tasks. This is the intended behavior for an Arize integration skill.- [EXTERNAL_DOWNLOADS]: The documentation provides standard installation instructions for the arize-ax-cli package from official package registries. This is expected and necessary for the tool's functionality.- [PROMPT_INJECTION]: While the skill ingests external span and experiment data to be used in evaluator prompts (representing a surface for indirect prompt injection), it includes critical instructions to never fabricate results and provides clear guidance on how to structure these prompts safely.

arize-evaluator