Investigate LLM failures with Phoenix traces.
What it does
Investigates LLM application failures and behavior by querying Phoenix traces, sessions, datasets, and experiments.
When to use it
Use when working with a Phoenix instance from the terminal or investigating and measuring its data with the CLI.
How to use it
Give it a Phoenix investigation or measurement request; it queries the instance and reports its findings.
Uses
Good to know
- starts 1 helper agents
- acts without asking
Latest review
What you need · 5
Requires the Phoenix CLI, either installed globally or invoked through npx with Node.js.
Requires a Phoenix instance to query or investigate.
Enables optional JSON processing of CLI output.
Access · 1
This skill
Phoenix
Write
Adds annotations in Phoenix