Arize Prompt Optimization

This is for teams running LLMs in production who want to make their prompts better based on real usage data. It connects to Arize's observability platform, pulls your production traces to find where prompts live (usually in llm.input_messages spans), grabs whatever performance signal you have like eval scores or human annotations, then runs an optimization loop using the ax CLI. The workflow is extract current prompt from spans, gather feedback data from evals or datasets, then iterate. If you're already logging OpenInference traces and have evaluation data flowing, this gives you a structured way to tune prompts with actual evidence instead of guessing. Requires the ax CLI and Arize credentials set up.

skills.sh installs: 3.4K
GitHub stars: 72
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add arize-ai/arize-skills --skill arize-prompt-optimization --agent claude-code

Installs into .claude/skills of the current project.

Files

Arize Prompt Optimization Skill

SPACE — --space flags accept a space name (e.g., my-workspace) or a base64 space ID (e.g., U3BhY2U6...). Confirm a known name with ax spaces get NAME; ax spaces list is for human browsing.

Related skills

arize-prompts: Create, version, and label prompts in Prompt Hub with ax prompts (JSON messages, providers, labels such as production). Use that skill when the artifact should live in Arize; use arize-prompt-optimization below to improve prompt text from traces, datasets, and experiments.

Concepts

Where Prompts Live in Trace Data

LLM applications emit spans following OpenInference semantic conventions. Prompts are stored in different span attributes depending on the span kind and instrumentation:

SKILL.md on GitHub