Better LLM prompts, grounded in real failures.
Improves LLM prompts using production traces, evaluations, and human feedback.
When to use it
Use when the user asks to optimize prompts, improve model responses, improve output quality, or tune system prompts.
Give it a prompt improvement request; it analyzes performance data and returns revised prompt messages with an explanation.
Requires the ax CLI with a configured Arize profile.
Requires authenticated Arize access to the relevant space; the profile setup instructions recommend a space-scoped service key.
This skill
Arize
Arize experiments
Creates Arize experiments
your LLM
Sends prompts and performance records