Find weak spots in your LLM evaluation pipeline.
What it does
Identifies weaknesses in an LLM evaluation pipeline and returns prioritized findings with concrete fixes.
When to use it
Use when inheriting an evaluation system, questioning its trustworthiness, or starting without evaluation infrastructure; not to build an evaluator from scratch.
How to use it
Give it available evaluation artifacts, locally or through a connected observability server; it returns an impact-ordered findings report with fixes.
What you provide
- Your files
What you need · 2
When evaluation artifacts are absent, directs the agent to start error analysis on real or generated traces; an installation requirement is unstated.
When no production data exists, directs the agent to generate test inputs before running the pipeline and analyzing its traces; an installation requirement is unstated.
Access · 1
This skill
observability MCP server
Reads evaluation traces and results