error-analysis
Installation
SKILL.md
Error Analysis
Error analysis means reading your system's outputs to find out how it fails. It is broad. This plugin splits the work into targeted skills. Your job here is small: find the row below that matches the user's situation, tell the user which skill you are loading and why, then load that skill and follow its workflow from start to finish instead of improvising your own version of it.
| Situation | Skill to load |
|---|---|
| Has traces, wants to find failure modes, no established taxonomy yet | error-discovery |
| Has an existing eval pipeline and wants to know if it can be trusted | eval-audit |
| Has a known failure mode and wants an LLM judge for it | write-judge-prompt |
| Has an LLM judge or evaluator and wants to check its quality | validate-evaluator |
| Has no traces to review yet | generate-synthetic-data, then error-discovery |
| Wants a custom annotation interface for some other labeling task | build-review-interface |
| Wants to evaluate a RAG pipeline | evaluate-rag |
Most requests that say "error analysis" with traces in hand mean error-discovery. This file holds only routing. When in doubt about which row fits, ask the user instead of guessing. The workflow lives in the targeted skill.